COPYRIGHTread.pudn.com/downloads163/doc/742245/imageenhanceinmatlab… · Web viewImages that are...

Image Enhancement System on MATLAB platform.

MANAGEMENT SUMMARY

Images that are captured sometimes are not in the exact quality and standard that

some parties might want it to be. For example, a blurred image caused by unwanted

vibration during the process of capturing the subject’s image. From this mistake and

unwanted condition that cause the quality of the image to be poor it will cause precious

time and money to be wasted. Sometimes, the documentary of the whole event can be

wasted just because the image was not in a proper quality for further processing. For

example, if images that are captured during the eclipse are not sharp and blur, the

documenter for the eclipse might have to wait for another eclipse to occur just to

capture images that they needed.

Before hand, I have did my study and research on the image enhancing

techniques in order to prepare myself for the tasks that are about to be completed in the

future. So in this project, a few image enhancements technique have been implemented

so that user will get the desired image criteria that they seek for. A few algorithms have

been fitted in to suit the desired function that the user seek for in order to enhance the

quality of the image itself.

Faculty of Information Science and TechnologyMultimedia University 1


CHAPTER 1:

INTRODUCTION

Most of the time, image represents an object that exists in the real world and the

subject is converted in a either two-dimensional or even three-dimensional artifact. This

image usually has the similar characteristic of the object that the image represented [1].

For example, if an image of hand is being captured from a hand subject, the result of the

image created has to be exactly similar to what it being captured. If the hand has 2

fingers sticking out while the rest of the fingers are held closely to the palm, the image

has to be exactly what the subject is, which is 2 fingers sticking out while the rest of the

fingers are held closely to the palm.

Image processing refers to computer manipulation towards an image which has

been converted to an electronical file. A certain set of algorithm will then be

implemented to this image file and the result will be based on what type of algorithm is

the image file being exposed to. Being different from image processing, computer

graphics image usually are used in creation of artificial images while image processing

merely taking a real subject and manipulate it using certain type of software.



Image enhancement is more than creating an image that looks better than it

should be but then, sometimes it brings a lot more benefits by using image enhancement

in certain industries. For example, a bone x ray image file of an patient. Image

enhancement techniques can be applied here in order to help out doctors to trace to the

tinniest, smallest, and most detailed image file so that the doctor can trace the

problematic area on the bone itself.

Not only that, image enhancement can be applied in forensic field to help them

discover even to the smallest clue just by enhancing the quality of the image that they

have. As we all know, forensic deals from the biggest pieces of evidence to the smallest

that can be. So, it will be hard for the forensic team without those images enhancing

software that will be helping them out throughout the whole process of collecting

evidence.

After evaluating most of the problems faced by the industries that deal with

images, I can say that by gathering all the image enhancing software into one complete

product will be a great leap forward for all the industries that deal with images

everyday. By having all the functions into one complete product, these industries would

not have to go through all the troubles just to produce or acquire the quality of image

that they wanted.



The main objective of this system is to have four different image enhancing

functions, which are the Intensity Transformation, Histogram Processing, Linear Spatial

Filtering and last but not least, Frequency Domain Filtering or also known as Fourier

Transformation. Equipped with these 4 four functions will allows the end user to

manipulate the images that they needed to.

It is a simple program where the user will load the image desired into the

program, and from that point, the user can edit the images just by using the four

functions provided in the system. Besides that, having all four functions that are mainly

used in image enhancing scenarios, end users do not have to travel up and down

anymore to purchase four different type of compact discs that consists four different

programs each. This action itself saves not only money, but also time and energy.



CHAPTER 2:

LITERATURE REVIEW

This project consists of four indispensable components to be detailed, namely;

Intensity transformation.

Histogram processing.

Linear spatial filtering

Frequency domain filtering (Fourier Transform)

2.1 Intensity Transformation

The additive system or primary colours which is the red, green and blue or better

known as the RGB system is not the only way we can approach to colour but we can

also use the intensity, hue, and saturation system or known as IHS. In fact, the IHS

defines colour better than the RGB system because the IHS represents colours as nearly

as what the human eye sees.

The IHS system is based on a colour sphere where the vertical axis represent

intensity, saturation is represented by the radius and hue is being represented by the

circumference.



Figure 1 – Intensity, hue and saturation colour co-ordinate system.

The vertical axis which represents the intensity of the image can be measured by

the intense brightness variations that ranges from black represent by the number 0 to

white which is represented by the number 255. This axis is only measuring the intensity

of black and no other colours.

Hue means the dominant wavelength of the colour itself. Like intensity, hue can

be measured by using numbers ranging from 0 at the midpoint of red tones and the

number will slowly increases by the direction of anti clock wise along the

circumference until it reaches back to 0, which by then it will be numbered as 255 for a

complete circular motion around the circumference.



The purity of colours are represented by the saturation. When the saturation is at

the middle of the colour sphere, which means being 0, it means the colour is completely

impure. As the value of saturation gets higher until the maximum number of 255, the

colour of the image gets purer and more intense in colour.

This method is based on the transformation of RGB multispectral channels into

HIS (Hue-Intensity-Saturation). The Intensity component is the most important

component and can be defined as follows:

In this method, we substitute the Intensity component by the panchromatic

image and then the inverse transformation (HIS to RGB) is done. To compare the

results, all these methods were used in addition the proposed method based on wavelet

decomposition [21].

2.2 Histogram Processing



Histogram processing refers to the method where it redistributes all the pixel

value within an image so that the pixels are roughly equal to the number of pixels to

each of the user-specified output gray scale classes. Histogram processing not only

reduces the contrast in the very light or dark parts in the image but also increases the

contrast at the highly populated range of brightness value [2]. This technique is known

to enhance an image with the most contrast compare to any image enhancing

techniques.

How do we determine this grey scale transformation function? Assume our grey

levels are continuous and have been normalized to lie between 0 and 1. We must find a

transformation T that maps grey values r in the input image F to grey values s = T(r) in

the transformed image [18].

It is assumed that

T is single valued and monotonically increasing, and

for .

The inverse transformation from s to r is given by

r = T-1(s).



If one takes the histogram for the input image and normalizes it so that the area

under the histogram is 1, we have a probability distribution for grey levels in the input

image Pr(r).

If we transform the input image to get s = T(r) what is the probability

distribution Ps(s) ?

From probability theory it turns out that

where r = T-1(s).

Consider the transformation

This is the cumulative distribution function of r. Using this definition of T we

see that the derivative of s with respect to r is

Substituting this back into the expression for Ps, we get



for all .Thus, Ps(s) is now a uniform distribution function,

which is what we want [18] .

Figure 2 – Original image and after Histogram Processing.



Figure 3 – Another example of original image that went through Histogram Processing.

2.3 Linear Spatial Filtering

Linear spatial filtering is another set of function used to enhance the appearance

of an image. This function is designed to highlight or suppress specific features in an

image based on their spatial frequency [3].

Rough texture area in an image has high spatial frequency while as the smooth

area usually composed of lower spatial frequency. Rough texture are can be defined as



changes in tone are rapid in a small area while the smooth area has lower colour

variation.

A filtering system works by having a window of pixel such as 3x3 or 5x5 to

move along the column and row of the image and along when the window is moving

row by row and column by column, a set of algorithm is applied to change the recent

value or the pixel into a new desired end value depending on the algorithm itself.

Low pass filter are generally use to smoothen the appearance of an image where

high pass filter does the opposite where it sharpens the appearance of an image.

Figure 4 – Low pass filter.



Figure 5 – High pass filter.

In the spatial domain we look for differences in the gray levels between pixels,

often as an approximation of the derivative of the image function [20]. From the image

function f(x,y) one can determine the vector gradient of the image:

(f(x,y) = ( f/ x, f/ y )

= arctan2( f/ x , f/ y ) yields the direction image.

( ( f/ x)2 + ( f/ y)2 ) results in the large image, often named gradient image.

In the simplest case, one discretely approximates the derivative and the gradient

of the image as:

f/ x = f[x,y] - f[x-1,y]

g( f[x,y] ) = | f/ x | + | f/ y |



2.4 Frequency domain filtering (Fourier Transform)

Just like the spatial linear filtering, Fourier Transform is another type of filtering

which involves a window of pixels moving along the row and column of the image and

changing the value of each pixel depending on the algorithm applied.

Fourier Transform decomposed an image into its very own sine and cosine

components. The output of the transformation is represented by a frequency domain

while the input of an image is represented by the spatial domain equivalent [5].

Figure 6 - Example of how the Fast Fourier Transform can "clarify" points.



Figure 7 - Faint blood prints can often only be enhanced with the Fast Fourier

Transform

Figure 8 - The Fast Fourier Transform can also remove background patterns.



As we are only concerned with digital images, we will restrict this discussion to

the Discrete Fourier Transform (DFT) [19].

The DFT is the sampled Fourier Transform and therefore does not contain all

frequencies forming an image, but only a set of samples which is large enough to fully

describe the spatial domain image. The number of frequencies corresponds to the

number of pixels in the spatial domain image, i.e. the image in the spatial and Fourier

domains are of the same size [19].

For a square image of size N×N, the two-dimensional DFT is given by:

where f(a,b) is the image in the spatial domain and the exponential term is the basis

function corresponding to each point F(k,l) in the Fourier space. The equation can be

interpreted as: the value of each point F(k,l) is obtained by multiplying the spatial image

with the corresponding base function and summing the result [19].

The basis functions are sine and cosine waves with increasing frequencies, i.e.

F(0,0) represents the DC-component of the image which corresponds to the average

brightness and F(N-1,N-1) represents the highest frequency [19].



In a similar way, the Fourier image can be re-transformed to the spatial domain.

The inverse Fourier transform is given by:

To obtain the result for the above equations, a double sum has to be calculated

for each image point. However, because the Fourier Transform is separable, it can be

written as

where

Using these two formulas, the spatial domain image is first transformed into an

intermediate image using N one-dimensional Fourier Transforms. This intermediate

image is then transformed into the final image, again using N one-dimensional Fourier

Transforms. Expressing the two-dimensional Fourier Transform in terms of a series of

2N one-dimensional transforms decreases the number of required computations [19].



Even with these computational savings, the ordinary one-dimensional DFT has

complexity. This can be reduced to if we employ the Fast Fourier

Transform (FFT) to compute the one-dimensional DFTs. This is a significant

improvement, in particular for large images. There are various forms of the FFT and

most of them restrict the size of the input image that may be transformed, often to

where n is an integer [19].



CHAPTER 3:

METHODOLOGY

In order to create a system that covers all of the functions like intensity

transformation, histogram processing, linear spatial filtering and frequency domain

filtering which is most likely known as the Fourier Transform, first of all I need to

know what are they and what they are capable of doing.

Online information has been playing a very vital role throughout the process of

understanding each of them. The Internet, being the biggest information source since it

is first launch has been guiding me throughout the whole project. The Internet has been

providing me with useful information that helps me out whenever I need specific data.

Besides that, some certain book has provided me a vital source of information

although it is kind of hard to retrieve a specific information or keyword.

MATLAB, being one of the most powerful software at this time has enabled me

to do some practice and understanding how things work. By slowly building the

foundation of handling MATLAB, I am sure that with the help of some other additional

information, this system can be created. Since MATLAB is one of the program widely



used for research purposes, it is likely to say that MATLAB will not only ease out most

of the task that will be introduced, but it will also benefits the programmer and the end

user of this program.



3.1 Structure Diagram


0.0

LOAD

1.1

IntensityTransformation

HistogramProcessing

1.2

LinearSpatial

Filtering

1.3

FourierTransform

1.4

2.0

SAVE


3.2 System Flowchart


Start

LOAD

Choice


HistogramProcessing

LinearSpatial

Filtering

FourierTransform

End

SAVE


3.3 Intensity Transformation Flowchart


Start

LOAD

Choice


Controlchoice

Red Green Blue

SAVE

End


3.4 Histogram Processing Flowchart


Start

LOAD

Choice

HistogramProcessing

Controlchoice

4

SAVE

End

8 16 32 64 128 256


3.5 Linear Spatial Filtering Flowchart


Start

LOAD

Choice

Linear SpatialFiltering

Controlchoice

Unsharp

SAVE

End

Average

Disk

Gaussian

Laplacian

Log

Motion

Prewitt

Sobel


3.6 Fourier Transform Flowchart


Start

LOAD

Choice

FourierTransform

SAVE

End


CHAPTER 4: IMPLEMENTATION PROCESS / RESULTS

4.1 PROPOSED SOLUTIONS

Image has played a vital role in almost every single field in the world. Not only

it helps industries to provide entertainment to human, it also helps saves many life and

images also can paint the future of our generations and world. By saying that, we can

see that images can be considered one of the most important aspects in our life, no

matter for professional or even personal usage.

Sometimes, by having an unclear image of something, it can jeopardize the

outcome of something that follows up. For example, an unclear image of patient’s bone

x-ray might cost the patients life and many more. So certain function is a must for

enhancing certain type of images. But the problem that many industries face out there is

that most of these image enhancing softwares are sold separately and it can be a tedious

job collecting all those software.

By compiling these entire images enhancing software into one complete

product, industries out there will not have to go through those tedious jobs just to get a

certain type of image enhancing software that they need. Not only can that, by having

all in one complete product, industries can experiment out new way of image enhancing



rather than the usual way of image enhancing that they normally use because right now

they have all the access to different type of image enhancing.

After talking about the problems faced in the industries today in this world,

image enhancing technique can be consider one of the most important techniques that

all industries must have in order to gain the quality of images that industries wanted. So,

I am going to compile all four image enhancing techniques into a complete product so

that industries that requires more than one image enhancing techniques does not need to

look for the separate software instead, the industries can get it all in one complete

software. Techniques like intensity transformation, histogram processing, linear spatial

filtering and frequency domain filtering also known as Fourier Transform will be

included into this software. This system will allow the user to load any images that the

user wants to enhance and from there, the user can choose any one of the four functions

that the images will go through. After choosing the technique wanted, the system will

result out the desired image.

4.2 Implementation

Analysis.

Identifying the requirement of the project.

o MATLAB as the base platform for the software to run.



o Data and algorithms that are needed to be used for each image

enhancement techniques.

o Basic handling of the platform to ensure that the product turn out to be in

a working condition.

Planning project process.

o Setting up timelines and well planned schedule so that it is not taken for

granted since time is limited and important.

Design.

Brainstorming ideas.

o Think of ideas to get a good grip of what the program should do and how

will it looks like.

o Layout of where each of the functions going to be place and how it is

going to work.

Prototyping.

o Prepare a few prototypes or models and to see which one is more user-

friendly or at least, easiest way for user to understand the program.

Implementation.

Tailoring the interface.

o Placing the buttons, graphics and other control panels into the interface.



Getting the buttons to work.

o Constructing algorithms and function codes for each of the button

available on its interface so it would work in the order it is supposed to.

Evaluation.

Project documentation.

o Preparing the documentation and proper user guide.

Testing the system.

o Get over some friends to test out the product to see if it is user-friendly

and to determined if it is easy to understand what the program does.

o Feedbacks are gathered so the system could be improvised.

Validating the system.

o The system is validated so it works perfectly and error-free.

4.3 Major Variations From Original Plan.

Since I only acquired a basic understanding of how the MATLAB platform

works, a few of the advance features are not able to be used due to the lack of

understanding.

Time constraint is a major challenge when developing the system. So the system

could not be improvised into an advance system which consists of more features

and functions.



4.4 The program.

Basic overview.

Figure 9 – Basic overlook of the interface.

This is the overlook of the interface that has been developed. As we can see

from the first glace, there are few button with a controls and two major axes.



The load button.

Figure 10 – The load button.

This is the load button. What is does is that when the user click, a dialog box

will appear, giving the user a freedom to go to desired directory and to load up the

desired image file.

Figure 11 – The load dialog box.



Behind the load button, are a few lines of codes that gives the ability for the load

button to do so.

Figure 12 – The load code.

This is the codes behind that load button. As you can see, it is named

‘load_pushbutton’. As you can see in the first box, it starts off with a line of code that

gives the command to the MATLAB to open a dialog box and choose out a certain type

of image files that the user can choose. In this line, I have limited the type that the

dialog going to show to only files that have the extension of .jpg or the user can choose

to view all files with the other options. When the specific file is chosen, the filename

will be assigned into a variable which is name ‘I’ as shown in the second box. Directly

after that, the file that is saved into that variable will be assigned to the axes that goes

by the name ‘axesBefore’. The line ‘imshow(I)’ basically gives out the command to

reveal the image that is saved into the ‘I’ variable. The ‘guidata(hObject,handles)’

basically update the file. All of these are shown in the third box.



So let’s choose a file and see what happens.

Figure 13 – Image being choose.



Figure 14 – The image appeared.

After the image has appeared, from this point onwards it is the point where

different type of functions can be applied to this specific image. If the wrong image has

been chosen, simply click on the load button and choose the correct image. The correct

image will overwrite the current image thus revealing the corrected image. Let’s start

off with the first function, which is the intensity transformation.



The Intensity Transformation.

Figure 15 – Intensity Transformation’s panel control.

This is the default value for the intensity transformation. As we can see, the

panel control is being build up with three control sliders which each one has been assign

with a specific colour module. If we were to adjust a specific slider value, the intensity

of the colour will change thus creating a new colour tone for the image. Let’s try out the

red control slider and leave everything else where it is set as default.



Figure 16 – The control red slider has been set to max.

Figure 17 – The result when red slider control has been set to max.

As we can see from here, the image’s red colour has been intensify to the max

and as the result, we got an image which the colour red has been maxed out. Here are

the codes.



Figure 18 – Slider red’s code.

As we can see here, the value of the slider which is determined by the position

of the slider bar will be passed to the variable ‘sliderRedValue’ as shown in the first

box.

Figure 19 – The intensity histogram code.



The codes shown in the first box shows that the value gathered from all the 3

different types of slider that have been set up in the interface. Let’s concentrate on the

red slider. So the value that has been saved into the ‘sliderRedValue’ will be call out at

this section of codes. The image is then being saved into the variable ‘I’ so that the

value can be implement into the image, as shown in the second box. After the codes has

been call out and its has done enhancing the image according to the slider value as in

box three, the image will then be display at the next axes which is named ‘axesAfter’,

which is being explained in box four. That is the main idea how the Intensity Histogram

button works.



The Histogram Processing.

Figure 20 – The Histogram Processing’s panel control.

This is how the histogram processing’s panel control looks like. As we can see,

it is being constructed by one button and one pop-up menu. In the pop-up menu, there a

few options for the user to choose from depending on the type of different histogram

value.

Figure 21 – The value has been set to 4.



As the value 4 has been chosen and the button push, the algorithm behind the

button will work its value and transform the image into desired output.

Figure 22 – The result of Histogram Processing when the value 4 is set.

The image will be change into a grayscale form first before the algorithm is

being applied. Depending on the value being set, the histogram will produced different

type of result. Below are a few example of histogram processing under different value.



Figure 23 – Result when the value is being set to 16.

Figure 24 – Result when the value is being set to 64.



Figure 25 – The pop-up menu in the Histogram Processing’s panel control.

In here, the switch case is being used. When a value is being chosen, it will then

be stored into a variable called ‘Value’.



Figure 26 – The histogram processing’s code.

Like as explained earlier, the image will first be transformed into grayscale form

with this function line, ‘I=rgb2gray(I)’ as in box one. The value from the pop-up menu

will then be call and store into the variable ‘histogramValue’ as in box two. In box

three, the histogram algorithm will be call out to edit the image based on the input

received in box two.



The Linear Spatial Filtering.

Figure 27 – The linear spatial filtering’s panel control.

In the linear spatial filtering control panel, the pop-up menu is being used once

again because it gives a clearer and better understanding to the user that the options

available are only in the predefined menu. Let’s see what the ‘unsharp’ filter does.



Figure 28 – The unsharp filter is being applied.

If we were to compare this result image to the original one, we can see that in

this image, the lines are being roughened creating a clear cut where the colour line is

being formed. Different type of filter gives out different types of result. Let’s see what

some of these filters do.



Figure 29 – The average filter is being applied.

Figure 30 – The disk filter is being applied.



Figure 31 – The motion filter is being applied.

As we can see from here, different filter gives out different result. From the wide

range of filters been provided, the user can choose any of these filters to suit the

purpose of his or her image enhancement desire.



Figure 32 – The linear spatial filtering’s code.

As the value from the pop-up menu is been retrieved from the pop-up menu

code as shown in box one, the value is then being passed to a variable which goes by

the name of ‘buttonText’. After the value has been stored, the code in the second box

will then be executed and that’s the main algorithm for the linear spatial filtering.



The Fourier Transformation.

Figure 33 – The fourier transform’s button.

Simple as it may seem, this is the fourier transformation button. Fourier

transformation is different from the other three functions being reviewed earlier. Fourier

transformation takes the image and converts it to a frequency domain first before

applying the algorithm. After that the modified frequency domain will be changed back

in spatial domain which is easier understood by our eyes, which is in image form. The

other three functions earlier directly apply the algorithm onto the spatial domain, in

other word the image, and produce a direct result.



Figure 34 – The result of fourier transform.

Besides that, when the fourier transform button is being pushed, another dialog

window will appear to show the user the image of the filter that is being used.

Figure 35 – The dialog box that contains the image of filter being used.



Figure 36 – The code for fourier transform.

In the red box, the image is retrieved from the first axes and then being stored

into the variable ‘f’. After that, it will go through a grayscale transformation. The

transformation into frequency domain and back in spatial domain is being down in the

area which is shown in box one. In the second box, the algorithm being used here is

called the Gaussian filtering. The third box shows the image of filter that is being used

for this transformation in a new dialog window.



The Save Button.

Figure 37 – The save button.

Like any other working program, when there is a load button, there surely to be

a save button. This save button works like normal save button which allow users to

choose the desired directory that the file will be save and what name will it be. When

the button is being clicked, a dialog box will appear so that the user can do the

necessary actions.

Figure 38 – The save dialog box.



Figure 39 – The save button’s code.

In the first box, we can see that the image shown at the second axes is being

retrieved and being assigned into a variable ‘I’. The first line in the blue box will make

the save dialog window appear so that the user can save the file in any directory desired.

The second line in the second box shows that the image from the second axes is being

saved into any file name that the user wished to name it.

Figure 40 – The desired file name is being input here.



CHAPTER 5:

CONCLUSION

This project is very tricky and challenging to tackle. Image enhancing

techniques that I’ve never heard before has to be applied. Even though creating a new

software a kind of easy but image enhancing techniques are practically new to me.

Thus far, the needed algorithm has been obtained and briefly understood.

Researched has took place and more detailed information has been gathered. The next

step will be implementing this algorithm in an interface which can be use by anyone out

there.

All in all, the progress of this subject has kicked off well enough that we can

complete the project in time.



REFERENCES

[1] Wikipedia: Definition of Image. Retrieved on November 30, 2007, from

http://en.wikipedia.org/wiki/Image

[2] The Integration of Remote Sensing Data into Geographic Information System.

Retrieved on December 1, 2007 from http://www.aubreyrhea.com/bookbag/geog553/

[3] Image Interpretation and Analysis – Image Enhancement. Retrieved on December 1,

2007 from http://ccrs.nrcan.gc.ca/resource/tutor/fundam/chapter4/05_e.php

[4] Fourier Transform. Retrieved on December 2, 2007 from

http://student.kuleuven.be/~m0216922/CG/fourier.html

[5] Image Transforms – Fourier Transform. Retrieved on December 2, 2007 from

http://homepages.inf.ed.ac.uk/rbf/HIPR2/fourier.htm

[6] CS6825: Computer Vision. Retrieved on December 3, 2007 from

http://www.mcs.csuhayward.edu/~grewe/CS6825/Mat/PointOperators/Equalization/

Equalization.htm


http://www.mcs.csuhayward.edu/~grewe/CS6825/Mat/PointOperators/Equalization/Equalization.htm

http://www.mcs.csuhayward.edu/~grewe/CS6825/Mat/PointOperators/Equalization/Equalization.htm


http://en.wikipedia.org/wiki/Image


[7] Point Operations – Histogram Equalization. Retrieved on December 3, 2007 from

http://www.cee.hw.ac.uk/hipr/html/histeq.html

[8] 55:148 Digital Image Processing. Retrieved on December 3, 2007 from

http://www.icaen.uiowa.edu/~dip/LECTURE/LinTransforms.html

[9] Sample of enhanced images. Retrieved on December 7, 2007 from

http://www.iisiltd.com/images.htm

[10]Sabrins Jr., (1987). Remote Sensing; principles and interpretations. New York: W.

H. Freeman

[11] MATLAB requirements. Retrieved on April 18, 2008 from

http://www.mathworks.com/products/matlab/requirements.html

[12] MATLAB 7.6 Download. Retrieved on April 18, 2008 from

http://www.mathworks.com/products/matlab/tryit.html

[13] MATLAB System Requirement – Windows. Retrieved on April 18, 2008 from

http://www.mathworks.com/support/sysreq/current_release/index.html



http://www.iisiltd.com/images.htm

http://www.icaen.uiowa.edu/~dip/LECTURE/LinTransforms.html

http://www.cee.hw.ac.uk/hipr/html/histeq.html


[14] MATLAB System Requirement – Linux. Retrieved on April 18, 2008 from

http://www.mathworks.com/support/sysreq/current_release/linux.html

[15] MATLAB System Requirement – Solaris. Retrieved on April 18, 2008 from

http://www.mathworks.com/support/sysreq/current_release/solaris.html

[16] MATLAB System Requirement – Mac. Retrieved on April 18, 2008 from

http://www.mathworks.com/support/sysreq/current_release/macintosh.html

[17] World Population – Wikipedia. Retrieved on April 18, 2008 from

http://en.wikipedia.org/wiki/World_population

[18] CS6825. Retrieved on April 21, 2008 from

http://www.mcs.csuhayward.edu/~grewe/CS6825/Mat/PointOperators/Equalization/

Equalization.htm

[19] Image Transforms – Fourier Transform. Retrieved on April 21, 2008 from


[20] RT2 Chapter 4 Image Enhancement. Retrieved on April 21, 2008 from

http://www.cs.ru.nl/~ths/rt2/col/h4/4verbeteringenENG.html



[21] Wavelet-based image fusion using “A trous” algorithm. Retrieved on April 21,

2008 from http://www.gisdevelopment.net/technology/ip/mi0353pf.htm



APPENDICES

Appendix 1: Business Plan

Section 1: The Business Profile

1.0 Project Risks

Since this project is a Image Enhancement System using MATLAB as the

platform, it require a firm knowledge when it comes to the coding section in MATLAB

language. The possible risks may appear on the software, operating system and browser.

First, the user computer needs to have the MATLAB software installed into the

computer in order to run the program. It is impossible for a computer without

MATLAB software to execute and run the Image Enhancement System because it is

in .m and .fig format, which is the format for MATLAB. To solve the problem, the

MATLAB has to be installed before hand. The installation file for the trial version can

be obtained from this website.


the lists of tables below show the minimum hardware requirement for each

operating system in order to install MATLAB. These tables taken from the MATLAB’s

website.






System requirement.

Windows

Operating System Processors Disk Space RAM

32-bit MathWorks™

Products

Windows® XP (Service Pack 1 or 2)

Windows Server 2003 (Service Pack 1 or 2, R2)

Windows Vista™

Intel® Pentium (Pentium 4 and above)

Intel Celeron

Intel Xeon

Intel Core

AMD Athlon™

AMD Opteron

AMD Sempron

510 MB (MATLAB® only)*

512 MB

(1024 MB recommended)

64-bit MathWorks™

Products

Windows® XP x64 (Service Pack 1 or 2)

Windows Server 2003 x64 (Service Pack 1 or 2, R2)

Windows Vista™


Intel Celeron

Intel Xeon

Intel Core

AMD64

510 MB (MATLAB® only)*

1024 MB


Table 1.1 – System requirement for Windows

* Disk space requirement varies depending on size of partition. The MATLAB installer

will inform you of the hard disk space requirement for your particular partition.

Installation size will be determined by the installer and can vary for NTFS and FAT

formats.



Graphics

16-, 24-, or 32-bit OpenGL capable graphics adapter

License Management

Some license types require a license server running FLEXnet 11.4, which is

provided by the MathWorks installer.

TCP/IP is required on all platforms when using a license server.

Additional Requirements

Microsoft Word 2002, 2003, or 2007 is required to run MATLAB® Notebook.

Microsoft Excel 2002, 2003, or 2007 is required to run MATLAB® Builder™

EX and Spreadsheet Link™ EX.



Linux


32-bit MathWorks™

Products

Red Hat Enterprise Linux v.4 and above

Fedora Core 4 and above

Debian 4.0 and above

Other distributions*


Intel Celeron

Intel Xeon

Intel Core

AMD Athlon**

AMD Opteron

AMD Sempron

510 MB (MATLAB® only)

512 MB


64-bit MathWorks™

Products

Debian 4.0 and above

OpenSuSE 10.1 and above

Other distributions*


Intel Celeron

Intel Xeon

Intel Core

AMD64


1024 MB


Table 1.2 – System requirement for Linux.

*Other Linux Distributions

For any other 32-bit or 64-bit Linux distribution not explicitly listed above, R2008a

requires the distribution to be built using Kernel 2.4.x or 2.6.x and glibc (glibc6) 2.3.4

and above. (glibc 2.3.4 users see above footnote).



**Processor must support SSE2 instruction set.

Graphics

16-, 24-, 32-bit OpenGL capable graphics adaptor

License Management






Solaris


64-bit MathWorks™

ProductsSolaris™ 10* ultraSPARC™


1024 MB


Table 1.3 – System requirement for Solaris.

* Running MATLAB R2008a on Solaris 10 requires that the OS have the current

recommended patches. For information on which patches to install please refer to

solutions 1-3UPLL6 and 1-431HK0 .

Graphics

24-bit graphics display for Sun Solaris™

License Management





http://www.mathworks.com/support/solutions/data/1-431HK0.html?solution=1-431HK0

http://www.mathworks.com/support/solutions/data/1-3NGJ7N.html?solution=1-3UPLL6


Macintosh


32-bit MathWorks™

Products

Mac® OS X 10.4 (10.4.8 and above)Mac OS X 10.5 (10.5.1 and above)

Intel 510 MB (MATLAB® only)

512 MB


Table 1.4 – System requirement for Macintosh.

Graphics

X11 (X server) for Mac® OS X

Java Virtual Machine

Java™ Runtime Environment 1.5 Release 4

License Management






Section 2: Commercialization

Market and Commercialization Potential

The main target of this project could be anyone who is interested in image

enhancing system. It could varies from students who is learning how each image

enhancing techniques work, to image editor who deals with images and needs a system

that can perform enhancing techniques, to forensic scientist who needs these techniques

to get a clearer images and many more.

In March 2008, the world's population is believed to have reached over 6.65

billion. In line with population projections, this figure continues to grow at rates that

were unprecedented before the 20th century, although the rate of increase has almost

halved since its peak, which was reached in 1963, of 2.2 percent per year. The world's

population, on its current growth trajectory, is expected to reach nearly 9 billion by the

year 2050. (http://en.wikipedia.org/wiki/World_population). Since our target market are

from student all the way to professional, a rough percentage of 40% will be taken from

the total world’s population, which gives out to the figure of 2.66 billion people. It

means to say that 2.66 billion people in this world can be considered the future end user

of this product. And by the year of 2050, by using the same percentage, a total number

of 3.6 billion people in this world will fall into that targeted market. That gives this

product a very high chance of ending up besides an end user’s computer.


http://en.wikipedia.org/wiki/World_population

http://en.wikipedia.org/wiki/Population

http://en.wikipedia.org/wiki/Population_growth

http://en.wikipedia.org/wiki/Population

http://en.wikipedia.org/wiki/March_2008


The first way to marketing the Image Enhacement System is through Compact

Discs (CDs). The system will be load into a CD just like any other softwares out there

that are sitting in the shelves. Besides that, demo CD can be handed out freely to almost

every colleges, university, workplace, laboratories, image studios any many other places

in order to create awareness that there is such system that allows user to get four mainly

and widely used image enhancement techniques in one CD.

Besides that, since the Internet is the way of life right now, websites can be set

up in order to create awareness and to promote the system. Setup file can also be

uploaded into the website so that user from all around the world can have access to the

system. The system is just a click away in these days.

It is possible for this system to get into shelves and the Internet. Although the

Internet seems to be the most profitable way of getting it widely spread across the

world, but having a system setup file in a compact disc is another way spreading the

system through users like a rapid fire.



Appendix 2: Meeting logs.


COPYRIGHTread.pudn.com/downloads163/doc/742245/imageenhanceinmatlab… · Web viewImages that are...

Documents

Transcript of COPYRIGHTread.pudn.com/downloads163/doc/742245/imageenhanceinmatlab… · Web viewImages that are...