Memoire Online - Implementation of edge detection for a digital image

This is to certify that the project work entitled» Implementation of edge detection for digital image» is a record of original property work done by:

in the partial fulfillment of the requirement for the award of Bachelor of Science in Electrical and Electronics Engineering Department (EEE) of Kigali Institute of Science and Technology during the academic year 2007.

This dissertation fulfills the requirement of the institute regulations relating to the nature and standards of works for the award of Bachelor of Science in Electronics and Telecommunication Engineering Degree.

We, HARELIMANA Joyeuse (GS20031000), MBARUBUKEYE Innocent (GS20031281), NYIRANGENDAHIMANA Stéphanie (GS20031373), declare that all the work presented in this report is original.

To the best of our knowledge, the same work has never presented in KIST or any other universities or institutions of high learning.

A reference though was done and was provided from other people's work therefore, we declare that this work is ours.

First we would like to thank the Almighty GOD for keeping us alive during our whole study.

We express our gratitude to every one who contributes in our studies, especially to our government, who gives us the opportunity of studding at high learning institution and KIST administration board who gave us financial support for this final year project.

We would like to thank all staffs of Electrical and Electronics Engineering Department in KIST and Estates Officer for their support and extra knowledge, which they gave us.

We would like to extend special thanks to our supervisor Mr. RUBESH Anand for giving up many ideas and essential hint to support my research to be conducted.

We express our deep gratitude to Mr. Kanyarubira J.Baptiste, Mr. Nsekanabo Emmanuel and Mr. Nyirimihigo Célestin for their greater support without them we will not finish this work.

We are especially grateful to our parents whose patience; encouragement and understanding were vital to the completion of this study.

Once again it is pleasure to acknowledge all people who support us directly and indirectly in completion of this project.

ABSTRACT

The world is growing up in information technology where the processing of images is required also. From the past until now, human being is seeking how to improve knowledge and updating system from oldest to the latest ones.

As memorial, in past, human used scripts and drawing for keeping images to be used in future but that oldest technology was so poor because it took along time to achieve it.

After that, analog pictures and image processing came on, whereby analog cameras were used. This was so fast and doest not require energy and forces for processing it.

But the worst of this analog technology is the storage of this image and no further processing which can be carried out on this. As solution, the digital image processing is widely used for overcoming a problem of storing and further processing that has been observed in analog image processing.

In digital image processing, different technologies along with computer are used for better image processing. This work of digital image processing, the focus is on Edge detection for digital images.

In the oldest research, the different technologies such as photometric and geometric models are used whereby some failures were experienced.

During our work, these failures and difficulties are overcoming by using MATLAB software where the edges of image are well and easier detected.

Figure 2.1: Fundamental steps in digital image processing..........................................6

Table 2..4.1: shows the two Roberts Cross kernels....................................................18

Table 2..4.2: shows the sobel kernel...........................................................................20

Table 2..4.3: shows Prewitt's mask............................................................................22

Table 3.1.a: shows product and comand used............................................................35

Table 3.1.b: shows Excel files to use with each Excel version..................................35

Table 3.1.c: shows finding information about matlab...............................................35

Table 3.1.d: shows conversion between different formats........................................47

Table 3.1: conversion between image types.............................................................48

Table 4.1: shws comparison of number of pixels between PNG, GIF, BMP file formats and JPEG file format.....................................................................................62

Table 4.2: shows the results based on variation of contrast and brightness...............64

1.1. GENERAL INTRODUCTION.

Edge detection in the human brain takes place automatically. It is an important concept both in the area of image processing and in the area of object recognition, Without being able to determine where the edge of an object fall, a machine would be able unable to determine many things about that object such as shape, volume, and area. Being able to recognize an object is a key step towards the development of artificial intelligence.

Edge detection is the process of localizing pixel intensity transitions. The edge detection have been used by object recognition, target tracking, segmentation, and etc. Therefore, the edge detection is one of the most important parts of image processing.

Edge detection can be used for numerous applications in which access to an article is to be restricted to a limited number of persons.

The edge detection offers several benefits including simplicity, convenience, security and accuracy.

The edge detection is an important in most image processing techniques. It can identify the area of image where large change in intensity occurs.

These changes are associated with the physical boundary or edge in the scene from which the image is derived.

There are many techniques used in edge detection, but most of them can be grouped into two categories:

The search-based methods detect edges by looking for maxima and minima in the first derivative of an image.

The zero crossing based methods: search for zero crossing in the 2^nd derivative of the image in the order to find an image.

The area where images are processed using proprietary software, for example Photoshop where the implementation details of any image processing algorithm are inaccessible. Some algorithms are quite complex and the result may be sensible to subtleties in the implementation. An image that has been extensively processed using proprietary software may well be challenged in court. This problem is solved by using MATLAB software which may not be as user friendly as an application like Photoshop; however, being a general purpose programming language it provides many important advantages for forensic image processing.

1.2 PROBLEM IDENTIFICATION.

As already started in the introduction, edge detection is used to detect an object (solid object). Not only the image detection, but also the intensity of gray-level of that object is noted.

There are some software used to process image and the implementation details of any image processing algorithm are inaccessible (lack of using programming language). Our project is focused on software used to detect edges of image employing mainly the MATLAB program for solving this problem.

1.3. SCOPE OF THE STUDY.

The areas of this work are in electronics and telecommunication engineering, which are very wide fields.

This work is intended to implement the edge detection for digital image, so that it may be carried out to a big contour (face) identification of an object (an image).

1.4. TECHNIQUES USED.

In order to accomplish this task, the researchers require the use of the following methods:

· The study guide utilizes an extensive conversation containing edge detection objectives.

· Library search was used to find background of digital image processing in this project.

General methods and techniques were used to collect for this project as primary as well as secondary data collection. Secondary data was collected from published materials related to the topic. These include: book class notes and online information by visiting some websites on the Internet.

1.5. OBJECTIVES OF THE PROJECT.

1.6. PROJECT OUTLINE.

The first is Introduction. It discusses the motivation for the work is being reported, provides a brief overview of each of the main chapters that the reader will encounter.

The second chapter is Literature review, which provides details about what other others have done and hence set a benchmark for the current project as well as to justify the use of specific solution techniques.

The third chapter is Research methodology. This chapter deals with the methodology and materials used.

The fourth chapter is Results and discussion. It deals with the analysis of results obtained from different file format of images used in this work.

The fifth chapter is conclusion and recommendation. It gives summary of the main findings also the problems limitation is discussed.

2.1 INTRODUCTION TO THE IMAGE.

2.1.0. DIGITAL IMAGE.

An image may be defined as a two-dimensional function, f (x,y), where x and y are special(plane) coordinates, and the amplitude of f at any pair of coordinates (x,y) is called intensity or grey- level of the image at that point.

When x, y and the amplitude values of f are all finite, discrete quantities, we call this image a digital image.

2.1.1. DIGITAL IMAGE PROCESSING

Image processing: is to perform a particular series of operation on that image, such a set of the field of digital image processing refers to processing digital images by means of a digital computer.

Note that a digital image is composed of a finite number of elements, each of which has a particular location and value.

These elements are referred to as picture elements, images elements, pels and pixels.

The area of an image analysis (also called image understanding) is in between image processing and computer vision.

There are no clear-cut boundaries in the continuum from image processing at one end to computer vision at the other. However one useful paradigm is to consider three types of computerized processes in this continuum: low-level, mid-level and high level processes.

Low-level process: involves primitive operation such as image processing to reduce noise, contrast enhancement and image sharpening. A low level process is characterised by the fact that both its inputs and outputs are images.

Mid-level processing on images: involves task such as segmentation (partitioning an image into regions or objects), description of those objects to reduce them to a form suitable for computer processing and classification (recognition) of individual objects.

A mid-level process is characterized by the fact that its input general is images, but its outputs are attributes extracted from those images. (Example: edges, contours, and the identity of individual objects).

Finally, high level processing involves making sense, of an ensemble of recognized objects, as in image analysis, and, at the far end of the continuum, performing the cognitive function normally associated with vision.

Based on the preceding comments, we see that a logical place of overlap between image processing and image analysis is the area of recognition of individual regions or objects in an image.

Thus, what we call in this book digital image processing encompasses processes whose inputs and outputs are images and, in addition encompasses processes that extract attributes from images, up to and including the recognition of individual object.

Digital Image Processing (DIP) is a multidisciplinary science that borrows principles from diverse fields such as optic, surface physique, visuals psychophysics, computer science, and mathematics. The main application of image processing include: astronomy, ultrasonic imaging, remote sensing, Video communication and microscopy, among innumerable others.

In digital image processing system, an image processor reads out image data from a window of an image buffer memory having N-columns and M- rows.

2.1.2. FUNDAMENTAL STEPS IN DIGITAL IMAGE PROCESSING

It is helpful to divide the material covered in image processing into the two broad

categories methods whose inputs and outputs are images, and methods whose inputs may be images, but whose outputs are attributing extracted from those images. This organization is summarized in fig.2.1 above.

The diagram does not imply that every process an applied to an image, but the intension is to convey an idea of all the methodologies that can be applied to images for different purposes and possibly with different objectives.

1. Image acquisition: note that acquisition could be as simple as being given an image that is already in digital form. Generally, image acquisition stage involves pre-processing such as scaling.

2. Image enhancement: is among the simplest and most appealing areas of digital image processing. Basically, the idea behind enhancement techniques is to bring out detail that is obscured, or simply to highlight certain features of interest in an image. A familiar example of enhancement is when we increase the contrast of an image because «it looks better».

3. Image restoration: is an area that also deals with improving the appearance of an image. However, unlike enhancement, which is subjective, image restoration is objective, in the sense that restoration techniques tend to be based on mathematical or probabilistic models of image degradation. Enhancement, on the other hand, is based on human subjective preferences regarding what constitutes a «good» enhancement result.

4. Colour image processing: is an area that has been gaining in importance because of the significant increase in the use of digital images over the Internet.

5. Wavelets: are the foundations for representing images in various degrees of resolution (approving). In particular, this material is used for image data compression and for pyramidal representation, in which images are subdivided successively into smaller regions.

6. Compression: as the name implies, deals with techniques for reducing the storage required saving an image, or the bandwidth required transmitting it. Image compression is familiar (perhaps in advertently) to most users of computers in the form of image file extension, such as the jpg file extension used in JPEG image compression standard.

7. Morphological processing: deals with tools for extracting image components that are useful in the representation and description of shape. The material in this processing begins a transition from processes that output images to processes that output image attributes.

8. Representation and description: almost always follow the output of a segmentation stage, which usually is raw pixels data, constituting either the boundary of a region (i.e. the set of pixels separating one image region from another) or all the points in the region itself. In either case, converting the data to a form suitable for computer processing is necessary. The first decision that must be made is whether the data should be represented as a boundary or as s complete region.

Boundary representation is appropriate when the focus is on external shape characteristics, such as corners, and inflection. Choosing a representation is only part of the solution for transforming raw data into a form suitable for subsequent computer processing.

Description, also called feature selection, deals with extracting attributes that result in some quantitative information of interest or are basic for differentiating one class of objects from another.

9. Recognition: is the process that assigns a label to an object based on its description. So far we have said nothing about the need for prior knowledge or about the interaction between the knowledge base and the processing modules in fig.2.1.

Knowledge about a problem domain is coded into an image processing system in the form of a knowledge database. This knowledge may be as simple as detailing regions of images where the information of interest is known to be located, thus limiting the search that has to be conducted in seeking that information. The knowledge base also can be quite complex, such as an interrelated list of all major possible defects in a materials inspection problem or an image database containing high-resolution satellite images of a region in connection with change-detection applications.

In addition to guiding the operation of each processing module, the knowledge base also controls the interaction between modules.

This distinction is made in fig.2.1 by the use of double-headed arrows between the processing modules and the knowledge base, as opposed to single-headed arrows linking the processing modules.

2.1.3. IMAGE SEGMENTATION

Image segmentation is the material in the previous section 2.1.2 began a transition from an image processing methods whose input and output are images, to methods in which the input are images, but the outputs are attributes extracted from those images. Segmentation is another major step in that direction.

Image segmentation algorithms general are based on two basic properties of intensity values: discontinuities and similarity. In the first category, the approach is to partition an image based on abrupt changes in intensity, such as edges in an image.

The principle approach in second category is based on partitioning an image into regions that are similar according to a set of predefined criteria.

Thresholding, region growing, and region splitting and merging are examples of methods in this category.

2.1.4. IMAGE REPRESENTATION

An image is stored as a matrix using standard Matlab matrix conventions. There are five basic types of images supported by Matlab:

2.2 DIGITAL COMPUTER

2.2.1. DEFINITION OF DIGITAL COMPUTER

A digital computer is a combination of digital devices and circuits that can perform a programmed sequence of operations with a minimum of human intervention.

The program is a set of coded instruction that is stored in the computer's internal memory along with all of the data that the program requires.

When the computer is commanded to execute the program, it performs the instructions in the order that they are stored in memory until the program is completed.

2.2.2. WORKING OF COMPUTERS.

Computers do not think, the computer programmer provides a program of instruction and data that specifies very detail of what to do, what to do it to, and when to do it.

The computer is simply a high-speed machine that can manipulate data, solve problems, and make decision, all under the control of the program.

If a programmer makes a mistake in the program or puts in the wrong data, the computer will produce wrong results. A popular saying in the computer field is garbage in/garbage out.

2.3. INTRODUCTION TO CAMERAS.

2.3.0. FILM CAMERAS

Film cameras are the cameras whose use films in recording images as photographs or as a moving picture.

A still film camera is made of three basic elements: an optical element (the lens), a chemical element (the film) and a mechanical element (the camera body itself). As we'll see, the only trick to photography is calibrating and combining these elements in such a way that they record a crisp, recognizable image.

There are many different ways of bringing everything together. In this article, we'll look at a manual single-lens-reflex (SLR) camera. This is a camera where the photographer sees exactly the same image that is exposed to the film and can adjust everything by turning dials and clicking buttons. Since it doesn't need any electricity to take a picture, a manual SLR camera provides an excellent illustration of the fundamental processes of photography.

The optical component of the camera is the lens. At its simplest, a lens is just a curved piece of glass or plastic. Its job is to take the beams of light bouncing off of an object and redirect them so they come together to form a real image an image that looks just like the scene in front of the lens.

2.3.1. DIGITAL CAMERAS

A digital camera is an electronic device used to capture and store photographs digitally, instead of using photographic film like conventional cameras, or recording images in an analog format to magnetic tape like many video cameras. Modern compact digital cameras are typically multifunctional, with some devices capable of recording sound and/or video as well as photographs.

Video cameras are classified as devices whose main purpose is to record moving images.

They generally include a microphone to record sound, and feature a small liquid crystal display to watch the video during taping and playback.

In addition, many Live-Preview Digital cameras have a "movie" mode, in which images are continuously acquired at a frame rate sufficient for video.

The term digital still camera (DSC) most commonly refers to the class of live-preview digital cameras, cameras that use an electronic screen as the principal means of framing and previewing before taking the photograph. All use either a charge-coupled device (CCD) or a CMOS image sensor to sense the light intensities across the focal plane.

Many modern live-preview cameras have a movie mode, and a growing number of camcorders can take still photographs.

Digital single-lens reflex cameras (DSLRs) are digital cameras based on film single-lens reflex cameras (SLRs), both types are characterized by the existence of a mirror and reflex system. See the main article on DSLRs for a detailed treatment of this category.

A rangefinder is a focusing mechanism once widely used on film cameras, but much less common in digital cameras. The term rangefinder alone is often used to mean a rangefinder camera, that is, a camera equipped with a rangefinder

This category includes very high end professional equipment that can be assembled from modular components (winders, grips, lenses, etc.) to suit particular purposes. Common makes include Hasselblad and Mamiya. They were developed for medium or large format film sizes, as these captured greater detail and could be enlarged more than 35mm.

A line-scan camera is a camera device containing a line-scan image sensor chip, and a focusing mechanism. These cameras are almost solely used in industrial settings to capture an image of a constant stream of moving material. Unlike video cameras, line-scan cameras use a single array of pixel sensors, instead of a matrix of them. Data coming from the line-scan camera has a frequency, where the camera scans a line, waits, and repeats. The data coming from the line-scan camera is commonly processed by a computer, to collect the one-dimensional line data and to create a two-dimensional image. The collected two-dimensional image data is then processed by image-processing methods for industrial purposes.

The resolution of a digital camera is often limited by the camera sensor (usually a charge-coupled device or CCD chip) that turns light into discrete signals, replacing the job of film in traditional photography. The sensor is made up of millions of "buckets" that collect charge in response to light. Generally, these buckets respond to only a narrow range of light wavelengths, due to a color filter over each. Each one of these buckets is called a pixel, and a demosaicing/interpolation algorithm is needed to turn the image with only one wavelength range per pixel into an RGB image where each pixel is three numbers to represent a complete color.

The one attribute most commonly compared on cameras is the pixel count. Due to the ever increasing sizes of sensors, the pixel count is into the millions, and using the SI prefix of mega- (which means 1 million) the pixel counts are given in megapixels. For example, an 8.0 megapixel camera has 8.0 million pixels.

The pixel count alone is commonly presumed to indicate the resolution of a camera, but this is a misconception. There are several other factors that impact a sensor's resolution. Some of these factors include sensor size, lens quality, and the organization of the pixels (for example, a monochrome camera without a Bayer filter mosaic has a higher resolution than a typical color camera). Many digital compact cameras are criticized for having excessive pixels, in that the sensors can be so small that the resolution of the sensor is greater than the lens could possibly deliver.

This digital camera is partly disassembled. The lens assembly (bottom right) is removed, but the sensor (top right) still captures a usable image, as seen on the LCD screen (bottom left).

Since the first digital backs were introduced, there have been three main methods of capturing the image, each based on the hardware configuration of the sensor and color filters.

The first method is often called single-shot, in reference to the number of times the camera's sensor is exposed to the light passing through the camera lens. Single-shot capture systems use either one CCD with a Bayer filter mosaic it, or three separate image sensors (one each for the primary additive colors red, green, and blue) which are exposed to the same image via a beam splitter.

The second method is referred to as multi-shot because the sensor is exposed to the image in a sequence of three or more openings of the lens aperture. There are several methods of application of the multi-shot technique. The most common originally was to use a single image sensor with three filters (once again red, green and blue) passed in front of the sensor in sequence to obtain the additive color information. Another multiple shot method utilized a single CCD with a Bayer filter but actually moved the physical location of the sensor chip on the focus plane of the lens to "stitch" together a higher resolution image than the CCD would allow otherwise. A third version combined the two methods without a Bayer filter on the chip.

The third method is called scanning because the sensor moves across the focal plane much like the sensor of a desktop scanner. Their linear or tri-linear sensors utilize only a single line of photosensors, or three lines for the three colors. In some cases, scanning is accomplished by rotating the whole camera; a digital rotating line camera offers images of very high total resolution.

The choice of method for a given capture is of course determined largely by the subject matter. It is usually inappropriate to attempt to capture a subject that moves with anything but a single-shot system. However, the higher color fidelity and larger file sizes and resolutions available with multi-shot and scanning backs make them attractive for commercial photographers working with stationary subjects and large-format photographs. CMOS-based single shot cameras are also somewhat common.[5]

Common formats for digital camera images are the Joint Photography Experts Group standard ( JPEG) and Tagged Image File Format ( TIFF).

They usually store images in one of two formats TIFF, which is uncompressed, and JPEG, which is compressed. Most cameras use the JPEG file format for storing pictures, and they sometimes offer quality settings (such as medium or high).

JPEG is a compression algorithm developed by people the format is named after, the Joint Photographic Experts Group.

JPEG is big selling point is that its compression factor stores the image on the hard drive in less bytes than the image is when it actually displays.

JPEG uses lossy compression (lossy meaning "with losses to quality"). Lossy means that some image quality is lost when the JPG data is compressed and saved, and this quality can never be recovered.

TIFF is the format of choice for archiving important images. TIFF is THE leading commercial and professional image standard. TIFF is the most universal and most widely supported format across all platforms, Mac, Windows, Unix. Data up to 48 bits is supported.

TIFF image files optionally use LZW lossless compression. Lossless means there is no quality loss due to compression. Lossless guarantees that you can always read back exactly what you thought you saved, bit-for-bit identical, without data corruption. This is a critical factor for archiving master copies of important images. Most image compression formats are lossless, with JPG and PCD files being the main exceptions

GIF was developed by CompuServe to show images online (in 1987 for 8 bit video boards, before JPG and 24 bit color was in use). GIF uses indexed color, which is limited to a palette of only 256 colors (next page). GIF was a great match for the old 8 bit 256 color video boards, but is inappropriate for today's 24 bit photo images.

PNG is not so popular yet, but it's appeal is growing as people discover what it can do. PNG was designed recently, with the experience advantage of knowing all that went before. The original purpose of PNG was to be a royalty-free GIF and LZW replacement. However PNG supports a large set of technical features. Compression in PNG is called the ZIP method.

A light source say a candle emits light in all directions. The rays of light all start at the same point the candle's flame and then are constantly diverging. A converging lens takes those rays and redirects them so they are all converging back to one point. At the point where the rays converge, you get a real image of the candle.

The digital camera is one of the most remarkable instances of this shift because it is so truly different from its predecessor. Conventional cameras depend entirely on chemical and mechanical processes you don't even need electricity to operate them. On the other hand, all digital cameras have a built-in computer, and all of them record images electronically.

The new approach has been enormously successful. Since film still provides better picture quality, digital cameras have not completely replaced conventional cameras. But, as digital imaging technology has improved, digital cameras have rapidly become more popular.

2.4. EDGE DETECTION

2.4.1 INTRODUCTION TO THE EDGE DETECTION

Edge detection is the process of localizing pixel intensity transitions. The edge detection

have been used by object recognition, target tracking, segmentation, and etc. Therefore, the edge detection is one of the most important parts of image processing.

There mainly exist several edge detection methods (Sobel, Prewitt, Roberts, Canny ). These methods have been proposed for detecting transitions in images. Early methods determined the best gradient operator to detect sharp intensity variations .

Commonly used method for detecting edges is to apply derivative operators on images.

Derivative based approaches can be categorized into two groups, namely first and second order derivative methods. First order derivative based techniques depend on computing the gradient several directions and combining the result of each gradient. The value of the gradient magnitude and orientation is estimated using two differentiation masks. In this work, Sobel which is an edge detection method is considered. Because of the simplicity and common uses, this method is preferred by the others methods in this work. The Sobel edge detector uses two masks, one vertical and one horizontal. These masks are generally used 3×3 matrices. Especially, the matrices which have 3×3 dimensions are used in matlab. The masks of the Sobel edge detection are extended to 5×5 dimensions, are constructed in this work. A matlab function, called as Sobel 5×5, is developed by using these new matrices. Matlab, which is a product of The Mathworks Company, contains has a lot of toolboxes. One of these toolboxes is image toolbox which has many functions and algorithms. Edge function which contains several detection methods (Sobel, Prewitt,Roberts, Canny, etc) is used by the user.

The image set, which consist of 8 images (256×256), is used to test Sobel 3×3 and

Edge detection is one of the techniques used for detecting the gray-level discontinuities.

It is the most common approach for detecting meaningful discontinuities in gray-level.[3]

Edge detection is an important concept, both in the area of image processing and in the area of object recognition. Without being able to determine where the edges of an object fall a machine would be unable to determine many things about that object such as shape, volume, area and so forth. Being able to recognize an object is a key step towards the development of artificial intelligence.

The goal of edge detection is to mark the points in a digital image at which the luminous intensity changes sharply. Edge detection of an image reduces significantly the amount of data and filters out information that may be regarded as less relevant, preserving the important structural properties of an image.

2.4.2 EDGE PROPERTIES.

Edges may be viewpoint dependent - these are edges that may change as the viewpoint changes, and typically reflect the geometry of the scene, objects occluding one another and so on, or may be viewpoint independent - these generally reflect properties of the viewed objects such as surface markings and surface shape.

Edge detection of an image reduces significantly the amount of data and filters out information that may be regarded as less relevant, preserving the important structural properties of an image. There are many techniques for edge detection, but most of them can be grouped into two categories: Search-based and zero-crossing based.

-The search-based methods detect edges by looking for maxima and minima in the first derivative of the image, usually local directional maxima of the gradient magnitude.

-The zero-crossing based methods search for zero crossings in the second derivative of the image in order to find edges.

2.4.3 DETECTING AN EDGE.

Taking an edge to be a change in intensity taking place over a number of pixels, edge detection algorithms generally compute a derivative of this intensity change.

Many edge-detection operators are based upon the 1^st derivative of the intensity - this gives us the intensity gradient of the original data. Using this information we can search an image for peaks in the intensity gradient.

For higher performance image processing, the 1st derivative can therefore be calculated (in 1D) by convolving the original data with a mask:

The magnitude of the first derivative can be used to detect the presence of an edge at a point in an image.

Some other edge-detection operators are based upon the 2nd derivative of the intensity. This is essentially the rate of change in intensity gradient.

Here, we can alternatively search for zero-crossings in the second derivative of the image gradient.

The signs of the 2^nd derivative can be used to determine whether an edge pixel lies on the dark or light side of an edge.[17]

Once we have calculated our derivative, the next stage is to apply a threshold, to determine where the result suggests an edge to be present. It is useful to be able to separate out the regions of the image corresponding to the objects in which we are interested, from the region of the image that correspond to background. Threshold often provides n easy and convenient way to perform this segmentation.

Thresholding is the simplest method of image segmentation on the basis of deferent intensities in the foreground and background regions of an image. Individual pixels in image are marked as «object» if their value is greater than some threshold value.(assuming an object to be bright than the background) and as» background «pixels if their values is less than threshold value. The key parameter in thresholding is the choice of a threshold.

Several different methods for choosing a threshold exist. But a common used compromiser method is thresholding with hysteresis. This method uses multiple thresholds to find edges. We begin by using the upper threshold to fid the start point. Once we have a start point, we trace the edge's path through the image pixel by pixel, marking an edge whenever we are above the lower threshold. We stop marking our edge only when the value falls below our low threshold.

2.4.4 EDGE DETECTION ALGORITHMS.

An edge, by definition, is that drastic change in value from one spot to the next either in an image or in one's own eyes. In the digital world, an edge is a set of connected pixels that lie on the boundary between two regions. Ideally, a digital edge would have one pixel colored a certain value and one of its neighboring pixels colored noticeably different. Since the contrast between the pixels is drastic, we label it an edge. Unfortunately, when dealing with digital pictures, most edges are blurred. This leads to edges having a gradual change of pixels levels from one object to the next thus causing the edge to be less defined and harder to distinguish. In this instance, one would have to devise an intuitive way to distinguish gray level transitions.

Edge detection is the process of determining where the boundaries of objects fall within an image. There are several algorithms that exist to date that are able detect edges. Roberts Cross was the first technique used and is thus the fasted and simplest to implement. Lastly, there is the Canny edge detection algorithm. This uses several steps to detect edges in an image. However, this algorithm can be slow at times due to its higher degree of complexity.

The Roberts Cross algorithm performs a two dimensional spatial gradient convolution on the image. The main idea is to bring out the horizontal and vertical edges individually and then to put them together for the resulting edge detection.

The two Roberts Cross kernels, shown above in Figure 1, are designed with the intention of bringing out the diagonal edges within the image. The Gx image will enunciate diagonals that run from the top-left to the bottom-right where as the Gy image will bring out edges that run top-right to bottom-left. The two individual images Gx and Gy are combined using the approximation equation

Roberts Cross is not quite as effective as the Sobel technique. It does bring out edges, but when compared the same image only using the Sobel algorithm, the amount of edges detected is poor. The Sobel approach seems to work the best.

In general, kernels are designed so that it consists of an odd number of rows and columns. This helps to define a distinct center in each kernel. Sampling pixels on each side of the center pixel leads to a better convolution image. This yields a better edge detection image produced in the outcome. The Sobel algorithm takes this account into effect when selecting its kernels.

estimate is vector sum of a pair of orthogonal vectors . Each orthogonal vector is a directional derivative estimate multiplied by a unit vector specifying the derivative's direction. The vector sum of these simple gradient estimates amounts to a vector sum of the 8 directional derivative vectors.

Sobel which is a popular edge detection method is considered in this work. There exists a function, edge.m which is in the image toolbox. In the edge function, the Sobel method uses the derivative approximation to find edges. Therefore, it returns edges at those points where the gradient of the considered image is maximum. The horizontal and vertical gradient matrices whose dimensions are 3×3 for the Sobel method has been generally used in the edge detection operations. In this work, a function is developed to find edges using the matrices whose dimensions are 5×5 in matlab.

Each direction of Sobel masks is applied to an image, and then two new images are created. One image shows the vertical response and the other shows the horizontal response. Two images combined into a single image. The purpose is to determine the existence and location of edges in a picture. This two image combination is explained that the square of created masks pixel estimate coincidence each other as coordinate are summed. Thus new image on which edge pixels are located obtained the value which is the squared of the above summation. The value of threshold in this above process is used to detect edge pixels.

An algorithm is developed to find edges using the new matrices and then, a matlab function, which is called as Sobel 5x5.m, is implemented in matlab. This matlab function requires a grayscale intensity image, two-dimensional array. The result which is returned by this function is the final image in which the edge pixels are denoted by white color.

The Sobel kernels, as shown in the above figure, are larger than its predecessor, the Roberts Cross, is. This causes the processes to be less susceptible to noise.

As one can see, there are only really two pixels of concern with the Roberts Cross algorithm thus noise can be a rather large issue. The Sobel algorithm deals with six pixels and in doing so takes a better average of the neighboring pixels. This along with slight blurring can eliminate most noise found within normal images.

First the image is run through a Gaussian blur to help get ride of a majority of the noise. Second, an edge detection algorithm is applied such as the Roberts Cross or Sobel techniques. From there the angle and magnitude can be obtained and used to determine which portions of the edges should be suppressed and which should be brought out. Lastly, there are two threshold cutoff points. Any value in the image below the first threshold is dropped to zero. Any value above the second threshold is raised to one. For any pixels whose values fall within the two thresholds, their values are set to zero or one, based on their neighboring pixels and their angle. The Canny algorithm first requires that the image be smoothed with a Gaussian mask, which cuts down significantly on the noise within the image. Then the image is run through the Sobel algorithm, and as discussed before, this process is hardly affected by noise. Lastly, the pixel values are chosen base on the angle of the magnitude of that pixel and its neighboring pixels. When using the canny edge detector there are two threshold levels. The first threshold is used to set all values in the image that fall below it to be set to zero. The second threshold level is used to set all values in the image that fall above it to be set to one. The values in the image that lay within the two threshold values are set to either zero or one based on their angle as specified by that algorithm. Thus, the image created is a binary image. Ideally, the lower threshold should be set relatively low so that a majority of the pixels are valued based on their angles. Then the upper threshold is set to a high value there tends to be a spurring effect that occurs on edges that branch off in a 'Y' shape. Unfortunately, when dealing with clouds or fog, this can be quite frequent and, as a result, there can be many spurs appearing in the result. Thus the value of the high threshold should not be set too had so it can, in a sense, prune the spurs away some.

Canny, which should be the best, might not be implemented correctly and therefore, does not work as it should. Currently it is able to detect edges that are well defined, but for minor edges, the best results are created when using the Sobel option. If the Canny detector worked properly it would be superior to both Sobel and Roberts Cross, the only drawback is that it takes longer to compute.[17]

The Prewitt edge detection is an alternative approach to the differential gradient edge detection. The operation usually output two images, one estimating the local edge gradient magnitude and other ones, estimating the orientation of the image.

The Prewitt operator is similarly to the sobel and some other operators approximating the first derivative, operators approximating the first derivatives are also called Compass operator because of the ability to determine gradient direction.

The gradient is estimated in eight (for 3X3 convolution mask) possible direction and convolution result of greatest magnitude indicate the gradient direction. Larger mask are possible.

Prewitt edge detection produces an image where higher grey-level values indicate the presence of an edge between two objects. The Prewitt edge detector mask is one of the oldest and best understand methods of detecting edges in images. Basically there are two masks, one for detecting image derivative in X and other one for detecting image derivative in Y.

In practice, one usually thresholds the result of Prewitt edge detection in other to produce a discrete set of edge.

2.5. PIXELS

In order for any digital computer processing to be carried out on an image, it must first be stored within the computer in a suitable form that can be manipulated by a computer program. The most practical way of doing this is to divide the image up into a collection of discrete (and usually small) cells, which are known as pixels. Most commonly, the image is divided up into a rectangular grid of pixels, so that each pixel is itself a small rectangle. Once this has been done, each pixel is given a pixel value that represents the colour of that pixel. It is assumed that the whole pixel is the same colour, and so any colour variation that did exist within the area of the pixel before the image was discretized is lost. However, if the area of each pixel is very small, then the discrete nature of the image is often not visible to the human eye. Other pixel shapes and formations can be used, most notably the hexagonal grid, in which each pixel is a small hexagon. This has some advantages in image processing, including the fact that pixel connectivity is less ambiguously defined than with a square grid, but hexagonal grids are not widely used.

Part of the reason is that many image capture systems (e.g. most CCD cameras and scanners) intrinsically discretize the captured image into a rectangular grid in the first instance.[14]

Each of the pixels that represent an image stored inside a computer has a pixel value, which describes how bright that pixel is, and/or what colour it should be. In the simplest case of binary images, the pixel value is a 1-bit number indicating either foreground or background. For a greyscale images, the pixel value is a single number that represents the brightness of the pixel. Often this number is stored as an 8-bit integer giving a range of possible values from 0 to 255. Typically zero is taken to be black, and 255 is taken to be white. Values in between make up the different shades of grey.

To represent colour images, separate red, green and blue components must be specified for each pixel (assuming an RGB colourspace), and so the pixel `value' is actually a vector of three numbers. Often the three different components are stored as three separate `greyscale' images known as colour planes (one for each of red, green and blue), which have to be recombined when displaying or processing.

Multi-spectral images can contain even more than three components for each pixel, and by extension these are stored in the same kind of way, as a vector pixel value, or as separate colour planes.[13]

The actual greyscale or colour component intensities for each pixel may not actually be stored explicitly. Often, all that is stored for each pixel is an index into a colourmap in which the actual intensity or colours can be looked up.

Although simple 8-bit integers or vectors of 8-bit integers are the most common sorts of pixel values used, some image formats support different types of value, for instance 32-bit signed integers or floating point values. Such values are extremely useful in image processing as they allow processing to be carried out on the image where the resulting pixel values are not necessarily 8-bit integers. If this approach is used then it is usually necessary to set up a colour map, which relates particular ranges of pixel values to particular displayed colours.

Brightness / Contrast: allows alteration of brightness and contrast of selected pixels or over the entire RGB image if no pixels have been selected.

Brightness: this is the amount of light intensity or received by the eye regardless of color.

Luminance: is the quantity of light intensity emitted per square centimetre of an illuminated area.

Saturation: this indicates the amount of all colours present in the given picture.

Brightness makes the image lighter or darker overall, while Contrast either emphasizes or de-emphasizes the difference between lighter and darker regions.

Contrast is easy to understand visually. Artistically, contrasting colors are colors that are opposite on the color wheel colors that are opposites. In a high contrast image, you can see definite edges and the different elements of that image are accented. In a low contrast image, all the colors are nearly the same and it's hard to make out detail.
Contrasting colors in terms of a computer's representation of an image, means the "primary colors" or the colors with color components of 0 or 255 (Min and Max). Black, White, Red, Green, Blue, Cyan, Magenta, and Yellow are the high contrast colours. When all the colors in an image are around one single color, that image has low contrast. Grey is the usual color of choice because it rests exactly in between 0 and 255 (127 or 128).

Describing colors using hue, saturation and brightness is a convenient way to organize differences in colors as perceived by humans. Even though color images on computer monitors are made up of varying amounts of Red, Green and Blue phosphor dots, it is at times more conceptually appropriate to discuss colors as made up of hue, saturation and brightness than as varying triplets of RGB numbers. This is because human perception sees colors in these ways and not as triplets of numbers.

If we imagine the three primary colors red, green and blue placed equally apart on a color wheel, all the other colors of the spectrum can be created by mixes between any two of the primary colors. For example, the printer's colors known as Magenta, Yellow, and Cyan are mid-way between Red and Blue, Red and Green and Blue and Green respectively.

This diagram is called the color wheel, and any particular spot on the wheel from 0 to 360 degrees is referred to as a hue, which specifies the specific tone of color. "Hue" differsslightly from "color" because a color can have saturation or brightness as well as a hue.

Saturation is the intensity of a hue from gray tone (no saturation) to pure, vivid color (high saturation).

Brightness is the relative lightness or darkness of a particular color, from black (no brightness) to white (full brightness). Brightness is also called Lightness in some contexts.

2.6. PRIMARY COLOURS

It is a useful fact that the huge variety of colours that can be perceived by humans can all be produced simply by adding together appropriate amounts of red, blue and green colours. These colours are known as the primary colours. Thus in most image processing applications, colours are represented by specifying separate intensity values for red, green and blue components. This representation is commonly referred to as RGB.

The primary colour phenomenon results from the fact that humans have three different sorts of colour receptors in their retinas which are each most sensitive to different visible light wavelengths.

The primary colours used in painting (red, yellow and blue) are different. When paints are mixed, the `addition' of a new colour paint actually subtracts wavelengths from the reflected visible light.

It is possible to construct (almost) all visible colours by combining the three primary colours red, green and blue, because the human eye has only three different colour receptor, each of them sensible to one of the three colours. Different combinations in the stimulation of the receptors enable the human eye to distinguish approximately 350,000 colours. A RGB colour image is a multi-spectral image with one band for each colour red, green and blue, thus producing a weighted combination of the three primary colours for each pixel.

A full 24-bit colour image contains one 8-bit value for each colour, thus being able to display different colours. However, it is computationally expensive and often not necessary to use the full 24-bit to store the colour for each pixel. Therefore, the colour for each pixel is often encoded in a single byte, resulting in a 8-bit colour image. The process of reducing the colour representation from 24-bits to 8-bits, known as colour quantization, restricts the number of possible colours to 256. However, there is normally no visible difference between a 24-colour image and the same image displayed with 8 bits. A 8-bit colour images are based on colourmaps, which are look up tables taking the 8 bit pixel value as index and providing an output value for each colour.

A colour perceived by the human eye can be defined by a linear combination of the three primary colours red, green and blue. These three colours form the basis for the RGB-colourspace. Hence, each perceivable colour can be defined by a vector in the 3-dimensional colourspace. The intensity is given by the length of the vector, and the actual colour by the two angles describing the orientation of the vector in the colourspace.

The RGB-space can also be transformed into other coordinate systems, which might be more useful for some applications. In this coordinate system, a colour is described with its intensity, hue (average wavelength) and saturation (the amount of white in the colour). This colour space makes it easier to directly derive the intensity and colour of perceived light and is therefore more likely to be used by human beings.

Full RGB colour requires that the intensities of three colour components be specified for each and every pixel. It is common for each component intensity to be stored as an 8-bit integer, and so each pixel requires 24 bits to completely and accurately specifies its colour. Image formats that store full 24 bits to describe the colour of each and every pixel are therefore known as 24-bit colour images.

Using 24 bits to encode colour information allows different colours to be represented, and this is sufficient to cover the full range of human colour perception fairly well.[14]

The term 24-bit is also used to describe monitor displays that use 24 bits per pixel in their display memories, and which are hence capable of displaying a full range of colours.

There are also some disadvantages to using 24-bit images. Perhaps the main one is that it requires three times as much memory, disk space and processing time to store and manipulate 24-bit colour images as compared to 8-bit colour images. In addition, there is often not much point in being able to store all those different colours if the final output device (e.g. screen or printer) can only actually produce a fraction of them. Since it is possible to use colourmaps to produce 8-bit colour images that look almost as good, at the time of writing 24-bit displays are relatively little used. However it is to be expected that as the technology becomes cheaper, their use in image processing will grow.

2.7. MULTI-SPECTRAL IMAGES

A multi-spectral image is a collection of several monochrome images of the same scene, each of them taken with a different sensor. Each image is referred to as a band. A well known multi-spectral (or multi-band image) is a RGB colour image, consisting of a red, a green and a blue image, each of them taken with a sensor sensitive to a different wavelength. In image processing, multi-spectral images are most commonly used for Remote Sensing applications. Satellites usually take several images from frequency bands in the visual and non-visual range. Landsat 5, for example, produces 7 band images with the wavelength of the bands being between 450 and 1250 nm.

All the standard single-band image processing operators can also be applied to multi-spectral images by processing each band separately. For example, a multi-spectral image can be edge detected by finding the edges in each band and than ORing the three edge images together. However, we would obtain more reliable edges, if we associate a pixel with an edge based on its properties in all three bands and not only in one.

To fully exploit the additional information which is contained in the multiple bands, we should consider the images as one multi-spectral image rather than as a set of monochrome greylevel images. For an image with k bands, we then can describe the brightness of each pixel as a point in a k-dimensional space represented by a vector of length k.

Special techniques exist to process multi-spectral images. For example, to classify a pixel as belonging to one particular region, its intensities in the different bands are said to form a feature vector describing its location in the k-dimensional feature space. The simplest way to define a class is to choose a upper and lower threshold for each band, thus producing a k-dimensional `hyper-cube' in the feature space. Only if the feature vector of a pixel points to a location within this cube, is the pixel classified as belonging to this class. A more sophisticated classification method is described in the corresponding worksheet.

The disadvantage of multi-spectral images is that, since we have to process additional data, the required computation time and memory increase significantly. However, since the speed of the hardware will increase and the costs for memory decrease in the future, it can be expected that multi-spectral images will become more important in many fields of computer vision.

2.8. LOOK UP TABLES AND COLOURMAPS

Look Up Tables or LUTs are fundamental to many aspects of image processing. An LUT is simply a table of cross-references linking index numbers to output values. The most common use is to determine the colours and intensity values with which a particular image will be displayed, and in this context the LUT is often called simply a colourmap.

The idea behind the colourmap is that instead of storing a definite colour for each pixel in an image, for instance in 24-bit RGB format, each pixel's value is instead treated as an index number into the colourmap. When the image is to be displayed or otherwise processed, the colourmap is used to look up the actual colours corresponding to each index number. Typically, the output values stored in the LUT would be RGB colour values.

There are two main advantages to doing things this way. Firstly, the index number can be made to use fewer bits than the output value in order to save storage space. For instance an 8-bit index number can be used to look up a 24-bit RGB colour value in the LUT. Since only the 8-bit index number needs to be stored for each pixel, such 8-bit colour images take up less space than a full 24-bit image of the same size. Of course the image can only contain 256 different colours (the number of entries in an 8-bit LUT), but this is sufficient for many applications and usually the observable image degradation is small.

Secondly the use of a colour table allows the user to experiment easily with different colour labeling schemes for an image.

One disadvantage of using a colourmap is that it introduces additional complexity into an image format. It is usually necessary for each image to carry around its own colourmap, and this LUT must be continually consulted whenever the image is displayed or processed.

Another problem is that in order to convert from a full colour image to (say) an 8-bit colour image using a colour image, it is usually necessary to throw away many of the original colours, a process known as colour quantization. This process is lossy, and hence the image quality is degraded during the quantization process.

Additionally, when performing further image processing on such images, it is frequently necessary to generate a new colourmap for the new images, which involves further colour quantization, and hence further image degradation.

As well as their use in colourmaps, LUTs are often used to remap the pixel values within an image. This is the basis of many common image processing point operations such as thresholding, gamma correction and contrast stretching. The process is often referred to as anamorphosis.

2.9. LUMINOUS INTENSITY

In photometry, luminous intensity is a measure of the wavelength-weighted power emitted by a light source in a particular direction, based on the luminosity function, a standardized model of the sensitivity of the human eye. The SI unit of luminous intensity is the candela (cd), an SI base unit. Photometry deals with the measurement of visible light as perceived by human eyes. The human eye can only see light in the visible spectrum and has different sensitivities to light of different wavelengths within the spectrum. When adapted for bright conditions (photopic vision), the eye is most sensitive to greenish-yellow light at 555 nm. Light with the same radiant intensity at other wavelengths has a lower luminous intensity. The curve which measures the response of the human eye to light is a defined standard, known as the luminosity function. This curve, denoted V(ë) , is based on an average of widely differing experimental data from scientists using different measurement techniques. For instance, the measured responses of the eye to violet light varied by a factor of ten.

Luminous intensity should not be confused with another photometric unit, luminous flux, which is the total perceived power emitted in all directions. Luminous intensity is the perceived power per unit solid angle. Luminous intensity is also not the same as the radiant intensity, the corresponding objective physical quantity used in the measurement science of radiometry.

One candela is defined as the luminous intensity of a monochromatic 540 THz light source that has a radiant intensity of 1/683 watts per steradian, or about 1.464 mW/sr. The 540 THz frequency corresponds to a wavelength of about 555 nm, which is green light near the peak of the eye's response. Since there are about 12.6 steradians in a sphere, the total radiant intensity would be about 18.40 mW, if the source emitted uniformly in all directions. A typical candle produces very roughly one candela of luminous intensity.

In 1881, Jules Violle proposed the Violle as a unit of luminous intensity, and it was notable as the first unit of light intensity that did not depend on the properties of a particular lamp. It was superseded by the candela in 1946.[15]

In an image processing context, the histogram of an image normally refers to a histogram of the pixel intensity values. This histogram is a graph showing the number of pixels in an image at each different intensity value found in that image. For an 8-bit greyscale image there are 256 different possible intensities, and so the histogram will graphically display 256 numbers showing the distribution of pixels amongst those greyscale values. Histograms can also be taken of colour images either individual histograms of red, green and blue channels can be taken, or a 3-D histogram can be produced, with the three axes representing the red, blue and green channels, and brightness at each point representing the pixel count. The exact output from the operation depends upon the implementation it may simply be a picture of the required histogram in a suitable image format, or it may be a data file of some sort representing the histogram statistics.[15]

The operation is very simple. The image is scanned in a single pass and a running count of the number of pixels found at each intensity value is kept. This is then used to construct a suitable histogram.

Binary images are images whose pixels have only two possible intensity values. They are normally displayed as black and white. Numerically, the two values are often 0 for black, and either 1 or 255 for white. Binary images are often produced by thresholding a greyscale or colour image, in order to separate an object in the image from the background. The colour of the object (usually white) is referred to as the foreground colour. The rest (usually black) is referred to as the background colour. However, depending on the image which is to be thresholded, this polarity might be inverted, in which case the object is displayed with 0 and the background is with a non-zero value.

Some morphological operators assume a certain polarity of the binary input image so that if we process an image with inverse polarity the operator will have the opposite effect. For example, if we apply a closing operator to a black text on white background, the text will be opened.

Intensity images measure the amount of light impinging on a photosensitive device. The input to the photosensitive device, typically a camera, is the incoming light, which enters the camera's lens and hits the image plane. In a digital camera, the physical image plane is an array which contains a rectangular grid of photosensors, each sensitive to light intensity. The output of the array is a continuous electric signal, the video signal. The video signal is sent to an electronic device called frame grabber, where it is digitised into a 2D rectangular array of integer values and stored in a memory buffer.

The interpretation of an intensity image depends strongly on the characteristics of the camera called the camera parameters. The parameters can be separated into extrinsic and intrinsic parameters. The extrinsic parameters transform the camera reference frame to the world reference frame. The intrinsic parameters describe the optical, geometric and digital characteristics of the camera. One parameter, for example, can describe the geometric distortion introduced by the optics.

2.10. GAUSSIAN SMOOTHING

The Gaussian smoothing operator is a 2-D convolution operator that is used to `blur' images and remove detail and noise. In this sense it is similar to the mean filter, but it uses a different kernel that represents the shape of a Gaussian (`bell-shaped') hump. This kernel has some special properties which are detailed below.

The idea of Gaussian smoothing is to use this 2-D distribution as a `point-spread' function, and this is achieved by convolution. Since the image is stored as a collection of discrete pixels we need to produce a discrete approximation to the Gaussian function before we can perform the convolution. In theory, the Gaussian distribution is non-zero everywhere, which would require an infinitely large convolution mask, but in practice it is effectively zero more than about three standard deviations from the mean, and so we can truncate the mask at this point.[10]

2.11. FACE IDENTIFICATION.

Face identification is very important in security, surveillance and telecommunication applications. The proposed algorithm will be used for face tracking in video surveillance system. In most cases the quality and resolution of the recorded image-segments is poor and hard to identify a face. For this reason, surveillance projects often use additional zooming cameras to focus on region of interests. Disturbing effects might distort recordings such as variations in illuminations, in pose or occlusions. Due to the acquisition conditions the size of the face images is smaller (sometimes much smaller) than the input image size in most of the existing face recognition systems. For example, the valid face region could be as small as 15×15 pixels, whereas the face images used in still image based systems are

at least 128 ×128 . Small-size images not only make the recognition task difficult, but also affect the accuracy of face segmentation.

In our approach we assume that the human head is detected and localized. The task is divided into the following sub-tasks. First some features are extracted to estimate the procession of the picture .Than the digital image is transformed to create edges on the boundary of digital image, in which the face is surrounded and scaled to a standard size. At last searching the most similar face in a database makes the identification of face.

This chapter deals with method used in conducting this project, the materials and equipments used, including origin and standard specifications.

3.1. METHOD USED AND SOFTWARE DEVELOPMENT.

In order to accomplish this task, general methods and techniques were used to collect data for this project as primary as well as secondary data collection.

Book class notes and online information by visiting some websites on the internet.

Library search was used to find background of digital image processing of our project.

The study guide utilizes an extensive conversation containing edge detection objectives. This conversation was done between researchers and supervisor; also it was between researchers and class mast.

After getting information about this project; we decided to use software called «MATLAB version 6.5»(MATLAB 6p5) to achieve our main objectives in this research.

3.1.1. DEFINITION OF MATLAB.

MATLAB (MATrix LABoratory) is a programming language for technical computing from the math works, Natick, MA. It is used for a wide variety of scientific and engineering calculation, especially for automatic control and signal processing. Matlab runs on windows, Mac and a variety of unit based systems Developed by Cleve Moler in the late of 1970's and based on the original LINPACK and EISPACK FORTRAN libraries, it was initially used for factoring matrices and solving linear equations. Moler commercialized the product with two colleagues in 1984. MATLAB is also noted for its extensive graphics capabilities.[11]

Double-click on the MATLAB icon (called a "shortcut") that the installer creates on your desktop.

Click on the Start button, select Programs, and click on the MATLAB 6.5 entry. Select MATLAB 6.5 from this menu.

Using Windows Explorer, open your top-level MATLAB installation directory and double-click on the shortcut to the MATLAB executable, MATLAB 6.5.

By default, when you start MATLAB using the shortcut the installer puts on your desktop, the initial current directory is the $MATLAB\work directory, where $MATLAB represents the name of your installation directory. The \work directory is a good place to store the M-files you modify and create because the MATLAB uninstaller does not delete the \work directory if it contains files. You can, however, use any directory as your MATLAB initial current directory. To specify another directory as your initial current directory, right-click on the MATLAB shortcut that the installer creates on your desktop and select the Properties option. Specify the name of the directory in the Start in field.

To include welcome messages, default definitions, or any MATLAB expressions that you want executed every time MATLAB is invoked, create a file named startup.m in the $MATLAB\toolbox\local directory. MATLAB executes this file each time it is invoked.

Certain products require additional configuration. The following table lists these products and the commands used to configure them. If you installed any of these products, see the documentation for that product for detailed configuration information.

By default, Excel Link (a separately orderable product) supports two versions of Excel. This table lists which Excel Link files to use with each Excel version. Use the appropriate files for your version of Excel. You can find these files in the $MATLAB\toolbox\exlink subdirectory.

After successfully installing MATLAB, you are probably eager to get started using it. This list provides pointers to sources of information and other features you may find helpful in getting started with MATLAB.

3.1.3. FUNCTION USED BY MATLAB.

1. graythresh function: is the function used to determine a threshold value.

2. strel function: is a function used to disappear the linear gaps if the sobel image is dilated using linear structuring elements.

3. imdilate function: is the function used to dilate the binary gradient mask using the vertical structuring element followed by the horizontal structuring element.

4. imfill function: is the function used to fill the interior holes(gaps) of image.

5. imclearboarder function: is used to remove the connected object to the border of the image. We create the diamond structuring element using the strel function.

6. bwperim function: is the function used to place an outline around the segmented object.

8. BW = edge (I,'sobel',thresh): specifies the sensitivity threshold for the Sobel method edge ignores all edges that are not stronger than thresh. If you do not specify thresh, or if thresh is empty ([]), edge chooses the value automatically.

MATLAB is a powerful programming language as well as an interactive computational environment. Files that contain code in the MATLAB language are called M-files. You create M-files using a text editor, then use them as you would any other MATLAB function or command.

Scripts, which do not accept input arguments or return output arguments. They operate on data in the workspace.

Functions, which can accept input arguments and return output arguments. Internal variables are local to the function.

If you're a new MATLAB programmer, just create the M-files that you want to try out in the current directory. As you develop more of your own M-files, you will want to organize them into other directories and personal toolboxes that you can add to your MATLAB search path.

SCRIPTS are the simplest kind of M-file because they have no input or output arguments. They are useful for automating series of MATLAB commands, such as computations that you have to perform repeatedly from the command line.

Scripts share the base workspace with your interactive MATLAB session and with other scripts. They operate on existing data in the workspace, or they can create new data on which to operate. Any variables that scripts create remain in the workspace after the script finishes so you can use them for further computations. You should be aware, though, that running a script can unintentionally overwrite data stored in the base workspace by commands entered at the MATLAB command prompt.

These statements calculate rho for several trigonometric functions of theta, then create a series of polar plots:

Try entering these commands in an M-file called petals.m. This file is now a MATLAB script. Typing petals at the MATLAB command line executes the statements in the script.

FUNCTIONS are program routines, usually implemented in M-files, that accept input arguments and return output arguments. They operate on variables within their own workspace. This workspace is separate from the workspace you access at the MATLAB command prompt.

Each M-file function has an area of memory, separate from the MATLAB base workspace, in which it operates. This area, called the function workspace, gives each function its own workspace context.

While using MATLAB, the only variables you can access are those in the calling context, be it the base workspace or that of another function. The variables that you pass to a function must be in the calling context, and the function returns its output arguments to the calling workspace context. You can, however, define variables as global variables explicitly, allowing more than one workspace context to access them. You can also evaluate any MATLAB statement using variables from either the base workspace or the workspace of the calling function using the evalin function.

The average function is a simple M-file that calculates the average of the elements in a vector:

Try entering these commands in an M-file called average.m. The average function accepts a single input argument and returns a single output argument. To call the average function, enter

The Primaty M-File Functions: is the first function in an M-file and typically contains the main program. Primary functios have a wide scope, the primary function can be invoked from outside of their M-File(from the matlab command line or from function in other file).

Subfunctions: act as subroutines to the main function. You can also use them to define multiple functions within a single M-file.

Nested Functions: are functions defined within another function. They can help to improve the readability of your program and also give you more flexible access to variables in the M-file.

Overloaded functions: are useful when you need to create a function that responds to different types of inputs accordingly. They are similar to overloaded functions in any object-oriented language.

3.1.4. COMMANDS USED.

0. Imread(): this command is used for reading an image. Within the parentheses, you type the name of an image file you wish to read. Put the filename within the single quotes.

1. Imshow(): this command is used to display an image represented as the matrix.

2. rgb2gray(): this command is used to convert between RGB format to intensity (grayscale) format.

3. figure: this command can be used to create a new current figure for the display. 4. Imdilate: this command is used to dilate the image.

5. figure(n): will check whether the figure n exists, if it switches the focus to it, otherwise a new window is created.

9. save x and load x: these commands are used to load and save variables in MATLAB. Once you have read a file, you probably convert it into an intensity image (a matrix) and work with this matrix; this matrix is represented by the variable x.

3.1.5. SYMBOLES USED.

1. ASTERISK (*): An asterisk in a filename specification is used as a wildcard specified.

2. COLON (:) the colon operator generates a sequence of numbers that you can use in creating or indexing into arrays.

5. DOT-DOT (..):two dots in sequence refer to the parent of the current directory.

6. EXCLAMATION POINT (!): The exclamation point precedes operating system commands that you want to execute from within MATLAB.

7. PARENTHESES ( ): Parentheses are used mostly for indexing into elements of an array or for specifying arguments passed to a called function.

8. PERCENT (%): The percent sign is most commonly used to indicate no executable text within the body of a program. This text is normally used to include comments in your code. Some functions also interpret the percent sign as a conversion specifier.

9. SEMICOLON (;): The semicolon can be used to construct arrays, suppress output from a MATLAB command, or to separate commands entered on the same line. Bt the way, semicolon prevents showing the result of the program line in workspace. This is handily especially with image.

10. SINGLE QUOTES (' '): Single quotes are the constructor symbol for MATLAB character arrays.

11. SPACE CHARACTER: The space character serves a purpose similar to the comma in that it can be used to separate row elements in an array constructor, or the values returned by a function.

12. SLASH AND BACKSLASH / \: the slash (/) and backslash (\) characters separate the elements of a path or directory string. On Windows-based systems, both slash and backslash have the same effect. On UNIX-based systems, you must use slash only.

13. SQUARE BLACKETS [ ]: Square brackets are used in array construction and concatenation, and also in declaring and capturing values returned by a function.

14. CURLY BRACES { }: Use curly braces to construct or get the contents of cell arrays,

3.1.6. TO WRITE MATLAB PROGRAMS

MATLAB programs can be written straight to command line (1), or in m-files. M-files are created by (2), or writing edit functionName on command line (1). Created files are stored as filename to current directory, which you can change in (3). You can, for example in case of a shared computer in laboratory exercises, create a temp directory under work for keeping your files organized.

Script is simple M-file that executes all lines as they were written in the command line. It has no «headers» or anything, just plain program lines. Scripts see all variables created in Command line (i.e. workspace) and vice versa.

Functions are «advanced scripts», that take can take parameters and can produce outputs. Examples of both a script and a function are given bellow:

To use a function, write [output1,outpu2,...] = functionName(param1, param2) on command line.

In case of single output, drop the []'s. MATLAB has numerous functions to help your work. You can find out how to use a function by typing help functionName on command line.

3.1.7. TO END A PROGRAM.

After writing a program; you have to end it or to finish it in other to run the writing program. To do this, you type «end» at the final line of your program.

3.1.8. ACCESS TO IMPLEMENTATION DETAILS

MATLAB provides many functions for image processing and other tasks. Most of these functions are written in the MATLAB language and are publicly readable as plain text files. Thus the implementation details of these functions are accessible and open to scrutiny. The defense can examine the processing used in complete detail, and any challenges raised can be responded to in an informed way by the prosecution. This makes MATLAB very different from applications, such as Photoshop.

It should be noted that some MATLAB functions couldn't be viewed. These are generally lower level functions that are computationally expensive and are hence provided as 'builtin' functions running as native code. These functions are heavily used and tested and can be relied on with considerable confidence.

3.1.9. FUNDAMENTALS ON MATLAB.

A digital image is composed of pixels, which can be thought of as small dots on the screen. A digital image is an instruction of how to colour each pixel. We will see in detail later on how this is done in practice. A typical size of an image is 512-by-512 pixels. Later on in the course you will see that it is convenient to let the dimensions of the image to be a power of 2. For example, 2⁹=512. In the general case we say that an image is of size m-by-n if it is composed of m pixels in the vertical direction and n pixels in the horizontal direction.

Let us say that we have an image on the format 512-by-1024 pixels. This means that the data for the image must contain information about 524288 pixels, which requires a lot of memory! Hence, compressing images is essential for efficient image processing.

MATLAB may not be as user friendly as an application like Photoshop, however, being a general purpose programming language it provides many important advantages for forensic image processing. It ensures the image processing steps used are completely documented, and hence can be replicated. In general, the source code for all image processing functions are accessible for scrutiny and test. It allows one to ensure numerical precision is maintained all the way through the enhancement process.

Image processing algorithms available under MATLAB are likely to be more advanced than those available from other image processing applications.

3.1.10. IMAGE TYPES AND DATA CLASSES

MATLAB supports intensity images (including binary images), indexed images, and color images. Most of this class will involve intensity images, in which each pixel is a scalar measure of light by a detector. Binary images have only values 0 or 1 for 'no light' or 'light'. Binary pixel values can be represented with one bit and packed into bytes.

Intensity images with more precision can use integer data of various lengths or floating point data of various lengths. Typically, we will use uint8 (unsigned 8-bit integers) or double (double-precision floating point numbers).

MATLAB has functions for converting the data types of images. It is important to understand the mathematics of these conversions.

Multiple pixels also can be indexed using a range. The first row is all pixels with row index 1 and any column index:

The keyword end can be used to indicate the last index. The range operator by itself, :, indicates the full range, from 1 to end either of a dimension or of the whole matrix.

MATLAB has standard arrays for creating arrays of a defined size with zeros, ones, true, false, or random values. For example:

MATLAB has many useful builtin functions listed in Appendix A of the text book. For example:

Note, given matrix p, max(p) treats the matrix as an array of column vectors and returns a vector of the largest value in each column.

Two images can be added so that the output value at each pixel is the sum of the values of the corresponding input pixels. For example:

The array and matrix arithmetic operations are done in double precision floating point. MATLAB also provides operations which support integer arithmetic, e.g., imadd and immultiply,

For multiplication and some other operations, MATLAB has two types of arithmetic operations: array arithmetic operations which perform the operation on corresponding pixels of the input images and matrix arithmetic operations which perform matrix arithmetic.

Most images you find on the Internet are JPEG-images which is the name for one of the most widely used compression standards for images. If you have stored an image you can usually see from the suffix what format it is stored in. For example, an image named myimage.jpg is stored in the JPEG format and we will see later on that we can load an image of this format into Matlab.

If an image is stored as a JPEG-image on your disc we first read it into Matlab. However, in order to start working with an image, for example perform a wavelet transform on the image, we must convert it into a different format. This section explains four common formats.

This is the equivalent to a "gray scale image" and this is the image we will mostly work with in this course. It represents an image as a matrix where every element has a value corresponding to how bright/dark the pixel at the corresponding position should be colored. There are two ways to represent the number that represents the brightness of the pixel: The double class (or data type). This assigns a floating number ("a number with decimals") between 0 and 1 to each pixel. The value 0 corresponds to black and the value 1 corresponds to white. The other class is called uint8, which assigns an integer between 0 and 255 to represent the brightness of a pixel. The value 0 corresponds to black and 255 to white. The class uint8 only requires roughly 1/8 of the storage compared to the class double. On the other hand, many mathematical functions can only be applied to the double class. We will see later how to convert between double and uint8.

This image format also stores an image as a matrix but can only color a pixel black or white (and nothing in between). It assigns a 0 for black and a 1 for white.

This is a practical way of representing color images. (In this course we will mostly work with gray scale images but once you have learned how to work with a gray scale image you will also know the principle how to work with color images.) An indexed image stores an image as two matrices. The first matrix has the same size as the image and one number for each pixel. The second matrix is called the color map and its size may be different from the image.

The numbers in the first matrix is an instruction of what number to use in the color map matrix.

This is another format for color images. It represents an image with three matrices of sizes matching the image format. Each matrix corresponds to one of the colors red, green or blue and gives an instruction of how much of each of these colors a certain pixel should use.

In some applications we want to study a sequence of images. This is very common in biological and medical imaging where you might study a sequence of slices of a cell. For these cases, the multiframe format is a convenient way of working with a sequence of images. In case you choose to work with biological imaging later on in this course, you may use this format.

3.1.11. CONVERSION BETWEEN DIFFERENT FORMATS

The following table shows how to convert between the different formats given above. All these commands require the Image processing toolbox!

The command mat2gray is useful if you have a matrix representing an image but the values representing the gray scale range between, let's say, 0 and 1000. The command mat2gray automatically re scales all entries so that they fall within 0 and 255 (if you use the uint8 class) or 0 and 1 (if you use the double class).

When you store an image, you should store it as a uint8 image since this requires far less memory than double. When you are processing an image (that is performing mathematical operations on an image) you should convert it into a double. Converting back and forth between these classes is easy.

3.2. EQUIPMENTS USED.

I n order to accomplish our project, we are used the equipments such as digital computer in processing of the object(image) and digital camera as the source of images. We can found more details of these equipments in section 2.2 and section 2.3.

3.3. MATERIALS USED.

In conducting this research, we were used materials of images like photos as samples.

3.4. DETECTING AN OBJECT USING IMAGE SEGMENTATION.

An object can be easily detected in an image if the object has sufficient contrast from the background. We use edge detection and basic morphology tools to detect a prostate cancer cell. To conduct this detection, there are many steps to follow:

%filename is the name of the object to be detected and .jpg is the file format of this image.

The image to be detected is the colour image, to detect it correctly; we have to convert it into gray scale (intensity value) using the command «rgb2gray».

We will detect this object; another word for object detection is segmentation. The object to be segmented differs greatly in contrast from the background image. Changes in contrast can be detected by operators; that calculate the gradient of an image. One way to calculate the gradient of an image is the Sobel operator, which creates a binary mask using a user-specified threshold value. We determine a threshold value using the graythresh function. To create the binary gradient mask, we use the edge function.

BWs = edge (I, 'sobel', (graythresh(I) * .1)); % detects edges of processed image.

figure, imshow(BWs), title('binary gradient mask'); %shows the binary gradient mask.

The binary gradient mask shows lines of high contrast in the image. These lines do not quite delineate the outline of the object of interest. Compared to the original image, you can see gaps in the lines surrounding the object in the gradient mask. These linear gaps will disappear if the Sobel image is dilated using linear structuring elements, which we can create with the strel function.

The binary gradient mask is dilated using the vertical structuring element followed by the horizontal structuring element. The imdilate function dilates the image.

The dilated gradient mask shows the outline of the cell quite nicely, but there are still holes in the interior of the cell. To fill these holes we use the imfill function.

(This step will be used when there are others object on the border of the detected object). The object of interest has been successfully segmented, but it is not the only object that has been found. Any objects that are connected to the border of the image can be removed using the imclearborder function. The connectivity in the imclearborder function was set to 4 to remove diagonal connections.

, in order to make the segmented object look natural, we smooth the object by eroding the image twice with a diamond-structuring element. We create the diamond-structuring element using the strel function.

An alternate method for displaying the segmented object would be to place an outline around the segmented object. The outline is created by the bwperim function.

3.5. IMPLEMENTATION OF EDGE DETECTION.

Sobel operators and Prewitt operators. These two operators are used in writing program, which were basically used in processing of digital images. Here is the whole program:

Writeable commands	Results obtained.
I=imread('dushime.jpg'); figure, imshow(I), title('original image');
rgb2gray('dushime.jpg'); newmap=rgb2gray(I); figure, imshow(newmap),title('gray image');
BWs=edge(newmap, 'sobel', (graythresh(newmap) * .1)); figure, imshow(BWs), title('binary gradient');
Se90=strel('line', 3, 90); Se0=strel('line', 3, 0); BWsdil=imdilate(BWs, [Se90 Se0]); figure, imshow(BWsdil), title('dilated gradient mask');
BWdfill=imfill(BWsdil,'holes'); figure, imshow(BWdfill), title('binary image with filled holes');
BWnobord=imclearborder(BWdfill,4); figure, imshow(BWnobord), title('cleared border image');
SeD=strel('diamond',1); BWfinal=imerode(BWnobord,SeD); BWfinal=imerode(BWfinal,SeD); figure, imshow(BWfinal), title('segmented image');
BWoutline=bwperim(BWfinal); segout=I; segout(BWoutline) = 255; figure, imshow(segout), title('outline original image'); end

The above program is used to process the JPEG format while the following program is used to process the BMP format.

I=imread('nouvelle.bmp'); figure, imshow(I), title('original image');
rgb2gray('nouvelle.bmp'); newmap=rgb2gray(I); figure, imshow(newmap),title('gray image');
BWs=edge(newmap, 'sobel', (graythresh(newmap) * .1)); figure, imshow(BWs), title('binary gradient');
Se90=strel('line', 3, 90); Se0=strel('line', 3, 0); BWsdil=imdilate(BWs, [Se90 Se0]); figure, imshow(BWsdil), title('dilated gradient mask');
BWdfill=imfill(BWsdil,'holes'); figure, imshow(BWdfill), title('binary image with filled holes');
BWnobord=imclearborder(BWdfill,4); figure, imshow(BWnobord), title('cleared border image');
SeD=strel('diamond',1); BWfinal=imerode(BWnobord,SeD); BWfinal=imerode(BWfinal,SeD); figure, imshow(BWfinal), title('segmented image');
BWoutline=bwperim(BWfinal); segout=I; segout(BWoutline) = 255; figure, imshow(segout), title('outline original image'); end

Writeable program

Obtainable result

% comparison for edge detection between Sobel & Prewitt operators.

filename= 'NEW.bmp'; % the filename to be read

image = imread(filename);

figure, Imshow(image), title ('original imge') % display

rgb2gray('NEW.bmp');

newmap=rgb2gray(image);

figure, imshow(newmap),title('gray image');

axis('square');

colormap('gray');

sze=size(newmap);

rows = sze(1); cols = sze(2);

% apply prewitt

sob_im = edge(newmap,'sobel');

figure, imshow(sob_im), title ('sobel image');

Figure 3.5.a. sobel image

prw_im= edge(newmap,'Prewitt');

figure, imshow(prw_im), title('prewitt image');

Figure 3.5.b. prewitt image.

dif_im=double(sob_im)- double(prw_im);

figure(2);

imagesc(dif_im);

axis('square');

colormap('gray');

end

4.1. RESULTS.

The basic requirement of this MATLAB program is to implement edge detection. The Sobel and Prewitt edge detector can also be applied to range the image (sampled picture) called Sobel figure and Prewitt figure. The corresponding edge image has been detected and can nice separated from the background.

The Sobel edge detection based on Sobel operator has the advantages of emphasizing the central part of the edge. It places emphasis on pixels that are closer to the center of the mask. Each direction of Sobel operator is applied to the image. One image shows the vertical response and other show the horizontal response. Purpose is to determine the existence and location of edge in picture.

When the object containing sharp edges and its surface is rather smooth, therefore could easily detect the boundary of the object without getting any erroneous pixels.

For the object containing many fine depth variations on its surface, applying the Sobel operator we can see that the intensity of many pixels on surface is high as long as the actual edges. Once reason is that the output of many edge pixels is greater than the maximum pixels values therefore there is `cutoff' at 255.

To avoid this over flow we scale the range image by a factor 0.25 prior to edge detection and then normalize the output although the result has improved significantly.

The Sobel operator is based on convolution in very small neighborhood and work well for specific image only.

The many disadvantages of this edge detection is its dependence on the size of object and sensitivity to noise.

The Sobel operator is very quick to computer and rather simple to implement. It yields the best result. The Sobel operator is one of the most commonly used edge detectors..

The Prewitt edge detector based on Prewitt operator approximates the first derivatives. This operator does not place any pixels that are closer to the center of the mask.

The Prewitt edge detector masks are one of the oldest and best understood methods of detecting edges in image. Basically there two mask one for detecting image derivatives in X and other one for detecting image derivatives in Y. The results of Prewitt edge detection are thresholded in order to produce a discrete set of edges. The direction of gradient mask is given by the mask giving maximal response.

The sobel and Prewitt operators use the first derivative. They have very simple calculation to detect edges and their orientation but they have inaccurate detection sensitivity in case of noise because the subjective evaluation of edge detection result images show that under noisy condition Prewitt and Sobel operator have poor quality.

As we have seeing in the precession of digital images (in chap.3); Prewitt and Sobel mark difference between them. The difference is that Sobel edge detector marks a lot of number of pixels while the Prewitt edge detector marks a few number of pixels.

The processed picture has been saved in different file format. The chosen file format can impact the clarity of picture. The following file formats have been chosen in this work such as JPEG, BMP, PNG, and GIF. The digital camera used when we shooted pictures offered JPEG setting, in order to achieve uor objective we tried to put the taken pictures into different file format using software that is able to convert them into the wanted format(e.g: Microsoft office picture manager).

The processed done using both sobel and Prewitt operators. A picture which is in JPEG file format when it is converted into other file format the number of pixels are variable according to the used operator(sobel or Prewitt).

When the picture is converted into GIF file format the sobel operators shows high number of pixels than JPEG, but Prewitt operator shows few number of pixels than the JPEG.

When the picture is converted into PNG file format, its procession with sobel shows high number of pixels than JPEG, but the Prewitt shows some number of number of pixels as JPEG. When a picture is converted into BMP file format, its procession with sobel and Prewitt operators show the same number of pixels as JPEG file format.

Table 4.1: shows comparison of number of pixels between PNG, GIF, BMP file formats and JPEG file format.

As we know that : the term contrast refer to the amount of color or gray scale differentiation that exist between various images features in both analog and digital image while brightness is the relative lightness or darkness of particular color from black to white. When a brightness of processed picture is increased at high level and contrast in the normal contrast of image's capturing, they were no pixels at the segmented image in both operator s(sobel and Prewitt).

When brightness is reduced at low level and contrast remains in the contrast of image's capturing, the unwanted pixels at segmented image occurred because of the background and the object were changed when the brightness was changed.

When contrast is increased at high level and brightness remains in normal condition, the Prewitt operator shows the unwanted pixels because there were no edges at segmented image and sobel operator shows us only the background.

When contrast was reduced at low level and brightness remains in its normal condition, pixels are available as the background remains unchangeable, there is no noise in both the sobel and Prewitt operators.

When contrast and brightness are increased there are no pixels on the face because the face tends to be as the background. But when both are reduced at low level there are the unwanted pixels on the segmented image, because pixels are located any where brightness is cleared.

Reducing contrast level and increasing brightness level no pixels are shown on segmented image in both sobel and Prewitt operators while reducing brightness level and increasing contrast level more pixels are shown on brig area only.

4.2. DISCUSSION.

The number of pixels in the picture depends on the resolution of used digital camera; small resolution corresponds to less number of pixels and high resolution results in more pixels.

The outlook of pixels on detected pictures depends on the weight of each picture. When a picture has small weight i.e. brightness and contast5 are at low level, pixels are not well seen. When a picture is being zoomed out its weight increase and pixels are well seen.

If the picture contains sharp edges and its surface is rather smooth, its boundaries could be detected easily without getting any erroneous pixels.

5.1. CONCLUSION.

As it was tried to express on the beginning of this paper, the main purpose of this project was intended to implement edge detection for digital image.

Our implementation was based on two edge detection operators which are Sobel and Prewitt operators, for determination of existence and location of edge in picture.

The Sobel operator yields the best result containing many pixels on the image. It is very quick to computer and rather simple to implement.

The Prewitt operator yields the result containing fewer pixels on the processed image.

Both operators pass two masks over the image separately to obtain X and Y magnitudes of the gradient

5.2. RECOMMENDATION.

Based on the investigation and experiences, we can give the following recommendations:

Fist one can extend this work by providing a smooth surface as background for the determination of the existence and location of edges in a picture.

Secondary, we would like to suggest usage of edge detection operators to improve security.

Thirdly, we would like to suggest using other techniques different of Sobel operator and Prewitt operators

To KIST administration, the period of final year project should be separated from the learning period.

[1]. Mathematical, Digital Image Processing. Powerful, Fast image processing and analysis.

[2]. Kristian Sandberg, Introduction to image processing in matlab 1. Department of applied mathematics, University of Colorado at Boulder.

[3]. Rafael C. Gonzalez, Richard E.Woods Digital Image Processing, Addison-Wesley Publishing Company, 1992.

[4]. Canny J. (1986), Computational approach to edge detection, IEEE Trans. Pattern Analysis and machine intelligence,p.679-714.

[5]. Ronald J. Tocci Neal S. widmer, Digital system, Principles and application, Eight edition.

[7]. Ma (1989),»A novel image sequence coding Schem», IEEE International conference on system engineering. p.15-18.

[10]. http://www.cee.hw.ac.uk/hipr/html/gsmooth.html[accessed 21^st July 2007] .

[12]. http:// en.wikipedia.org/wiki/image: partly disassembled Lunix digital camera. [Accessed 14^thseptember 2007].

[14]. http://www.freepatents on line.com/2003 0223615 [accessed 12^th august 2007].

[16]. (http://w ww.generation5.org/content/2002/im01.asp) [accessed 15^th October 2007]

SI photometry units
Quantity	Symbol	SI unit	Abbr.	Notes
Luminous energy	Q_v	lumensecond	lm·s	units are sometimes called talbots
Luminous flux	F	lumen (= cd·sr)	lm	also called luminous power
Luminous intensity	I_v	candela (= lm/sr)	cd	an SI base unit
Luminance	L_v	candela per square metre	cd/m²	units are sometimes called nits
Illuminance	E_v	lux (= lm/m²)	lx	Used for light incident on a surface
Luminous emittance	M_v	lux (= lm/m²)	lx	Used for light emitted from a surface
Luminous efficacy		lumen per watt	lm/W	ratio of luminous flux to radiant flux; maximum possible is 683.002

The tables below list all functions in the Image Processing Toolbox by category. The tables include a few functions in MATLAB that are especially useful for image processing, such as imagesc, immovie, and imwshow.

Image Display

colorbar

Display colorbar. (This is a MATLAB function. See the online MATLAB Function Reference for its reference page.)

getimage

Get image data from axes

image

Create and display image object. (This is a MATLAB function. See the online MATLAB Function Reference for its reference page.)

imagesc

Scale data and display as image. (This is a MATLAB function. See the online MATLAB Function Reference for its reference page.)

immovie

Make movie from multiframe indexed image

imshow

Display image

montage

Display multiple image frames as rectangular montage

subimage

Display multiple images in single figure

truesize

Adjust display size of image

warp

Display image as texture-mapped surface

zoom

Zoom in and out of image or 2-D plot. (This is a MATLAB function. See the online MATLAB Function Reference for its reference page.)

Neighborhood and Block Processing

bestblk

Choose block size for block processing

blkproc

Implement distinct block processing for image

col2im

Rearrange matrix columns into blocks

colfilt

Perform neighborhood operations using columnwise functions

im2col

Rearrange image blocks into columns

nlfilter

Perform general sliding-neighborhood operations

Image Analysis

edge

Find edges in intensity image

qtdecomp

Perform quadtree decomposition

qtgetblk

Get block values in quadtree decomposition

qtsetblk

Set block values in quadtree decomposition

Colormap Manipulation

brighten

Brighten or darken colormap. (This is a MATLAB function. See the online MATLAB Function Reference for its reference page.)

cmpermute

Rearrange colors in colormap

cmunique

Find unique colormap colors and corresponding image

colormap

Set or get color lookup table. (This is a MATLAB function. See the online MATLAB Function Reference for its reference page.)

imapprox

Approximate indexed image by one with fewer colors

rgbplot

Plot RGB colormap components. (This is a MATLAB function. See the online MATLAB Function Reference for its reference page.)

Morphological Operations (Binary Images)

applylut

Perform neighborhood operations using lookup tables

bwarea

Area of objects in binary image

bwareaopen

Binary area open; remove small objects

bwdist

Distance transform

bweuler

Euler number of binary image

bwfill

Fill background regions in binary image

bwhitmiss

Binary hit-miss operation

bwlabel

Label connected components in 2-D binary image

bwlabeln

Label connected components in N-D binary image.

bwmorph

Perform morphological operations on binary image

bwpack

Pack binary image

bwperim

Find perimeter of objects in binary image

bwselect

Select objects in binary image

bwulterode

Ultimate erosion

bwunpack

Unpack a packed binary image

imregionalmin

Regional minima of image

imtophat

Perform tophat filtering

makelut

Construct lookup table for use with applylut

Structuring Element (STREL) Creation and Manipulation

getheight

Get the height of a structuring element

getneighbors

Get structuring element neighbor locations and heights

getnhood

Get structuring element neighborhood

getsequence

Extract sequence of decomposed structuring elements

isflat

Return true for flat structuring element

reflect

Reflect structuring element

strel

Create morphological structuring element

translate

Translate structuring element

Pixel Values and Statistics

corr2

Compute 2-D correlation coefficient

imcontour

Create contour plot of image data

imfeature

Compute feature measurements for image regions

imhist

Display histogram of image data

impixel

Determine pixel color values

improfile

Compute pixel-value cross-sections along line segments

mean2

Compute mean of matrix elements

pixval

Display information about image pixels

regionprops

Measure properties of image regions

std2

Compute standard deviation of matrix elements

Toolbox Preferences

iptgetpref

Get value of Image Processing Toolbox preference

iptsetpref

Set value of Image Processing Toolbox preference