IMAGE COMPRESSION

1. INTRODUCTION

1.1 Project Overview Image compression is minimizing the size in bytes of a graphics file without degrading the quality of the image to an unacceptable level. The reduction in file size allows more images to be stored in a given amount of disk or memory space. It also reduces the time required for images to be sent over the Internet or downloaded from Web pages. There are several different ways in which image files can be compressed. For Internet use, the two most common compressed graphic image formats are the JPEG format and the GIF format. The JPEG method is more often used for photographs, while other techniques for image compression include the use of fractals and wavelets. These methods have not gained widespread acceptance for use on the Internet as of this writing. However, both methods offer promise because they offer higher compression ratios than the JPEG or GIF methods for some types of images. Another new method that may in time replace the GIF format is the PNG format. Compression is two types, one is Lossy Compression and another one is Loss Less Compression. Compression can produce an approximation of the image, in which case it is not possible to decompress the image and retrieve the original form is called lossy compression. Lossless and lossy compression are terms that describe whether or not, in the compression of a file, all original data can be recovered when the file is uncompressed.

IMAGE COMPRESSION

With lossless compression, every single bit of data that was originally in the file remains after the file is uncompressed. All of the information is completely restored. This is generally the technique of choice for text or spreadsheet files, where losing words or financial data could pose a problem; method is commonly used for line art and other images in which geometric shapes are relatively simple. Introduction to Image Compression In this project we use two methods of image compression. Huffman encoding. Haar Discrete Wavelet Transform algorithm. These two methods are lossless compression technique. In this project, we will use transform coding. Transform coding is an image compression technique that first switches to the frequency domain, then does its compressing. Here we use the Haar Discrete Wavelet Transform. The Haar transform operates as a square matrix of length

N = some integral power of 2.

In computer science and information theory, Huffman coding is an entropy encoding algorithm used for lossless data compression. The term refers to the use of a variable-length code table for encoding a source symbol (such as a character in a file) where the variable-length code table has been derived in a particular way based on the estimated probability of occurrence for each possible value of the source symbol.

IMAGE COMPRESSION 1.2. Definitions Image Images may be two-dimensional, such as a photograph, screen display, and as well as a three dimensional, such as a statue. They may be captured by optical devices such as cameras, mirrors, lenses, telescopes, microscopes, etc. and natural objects and phenomena, such as the human eye or water surfaces. Compression

Compression is the technique for reducing the storage required to save an image or the band width required to transmit it. Lossy Compression Lossy Compression is a data compression method which discards (loses) some of the data, in order to achieve its goal, with the result that decompressing the data yields content that is different from the original, though similar enough to be useful in some way. Lossy compression is most commonly used to compress multimedia data (audio, video, still images), especially in applications such as streaming media and internet telephony Lossless Compression Lossless data compression is a class of data compression algorithms that allows the exact original data to be reconstructed from the compressed data. The term lossless is in contrast to lossy data compression, which only allows an approximation of the original data to be reconstructed in exchange for better compression rates.

IMAGE COMPRESSION Huffman Code

In computer science and information theory, Huffman coding is an entropy encoding algorithm used for lossless data compression. Entropy Entropy is a lossless data compression scheme that is independent of the specific characteristics of the medium. One of the main types of entropy coding creates and assigns a unique prefix code to each unique symbol that occurs in the input. Encoding Encoding is an algorithm used for lossless data compression. The term refers to the use of a variable-length code table for encoding a source symbol (such as a character in a file) where the variable-length code table has been derived in a particular way based on the estimated probability of occurrence for each possible value of the source symbol. Decoding Decoding is a process of analog-to-digital conversion and digital-to-analog conversion. In this sense, these terms can apply to any form of data, including text, images, audio, video, multimedia, computer programs, or signals in sensors, telemetry, and control systems. Wavelet Compression Wavelet compression is a form of data compression well suited for image compression . The goal is to store image data in as little space as possible in a file. Wavelet compression can be either lossless or lossy.

IMAGE COMPRESSION Hamming

Hamming is the input of histogram bytes in the original file. Using this information, we can build the trees which allow us to calculate the variable length code of the compressed data. This class is simply an implementation of the algorithm we learnt in class. Predictive Coding In predictive coding, information already sent or available is used to predict future values, and the difference is coded. Since this is done in the image or spatial domain, it is relatively simple to implement and is readily adapted to local image characteristics. Transform Coding Transform coding, on the other hand, first transforms the image from its spatial domain representation to a different type of representation using some well-known transform and then codes the transformed values (coefficients).This method provides greater data compression compared to predictive methods, although at the expense of greater computation. Quantization The Quantization function, quant, is also called in squishier. We used a number of different bit allocation masks in order to determine which scheme is better. Mean Square Error Mean Square Error (MSE) of an estimator is one of many ways to quantify the difference between an estimator and the true value of the quantity being estimated.

IMAGE COMPRESSION Pixel

In digital imaging, a Pixel (picture element) is a single point in a raster image. The pixel is the smallest addressable screen element; it is the smallest unit of picture which can be controlled. Each pixel has its own address. Joint Photographic Experts Group Joint Photographic Experts Group (JPEG) is a commonly used method of lossy Compression for photographic images. The degree of compression can be adjusted, allowing a selectable tradeoff between storage size and image quality Peak Signal-to-Noise Ratio The phrase peak signal-to-noise ratio (PSNR) is most commonly used as a measure of quality of reconstruction of lossy, lossless compression. Mean Opinion Score The Mean Opinion Score (MOS) provides a numerical indication of the perceived quality of received media after compression and/or transmission. Differential Pulse Code Modulation Differential Pulse-code modulation (PCM) is a digital representation of an analog signal where the magnitude of the signal is sampled regularly at uniform intervals, then quantized to a series of symbols in a numeric (usually binary) code

IMAGE COMPRESSION

2. SYSTEM ANALYSIS
System analysis is the process of gathering and interpreting facts, diagnosing problems and using the information to recommend improvements to the systems. Analysis specifies what the system should do. The first stage of the software development is the study of the system under consideration and its requirement analysis. During these stages constant interaction with the users of the system is necessary. System analysis is the key of life cycle. To ensure the success of the system, careful and extensive analysis is required. Analysis is a study of various operations performed by the system. This involves gathering information and using structured tools for analysis. 2.1 Existing System There are several Image Compression software is available. Existing image compression algorithms (eg. ITU Group 4 and JBIG) offer efficient solutions to the storage problem but do not sufficiently support other objectives such as spatial access and fast decoding. 2.2 Drawbacks of Existing System Quality Loss Some problem arises while storing and transferring the image Decompress makes slight difference to the original image. Low compression ratio

IMAGE COMPRESSION 2.3 Proposed System

We propose a novel method based on JBIG, in which the other objectives are also met. The compression performance of the proposed method is only 10% worse than that of JBIG, and at the same time, spatial access to a compressed file is achieved. The method is also 2.5 times faster in decompression than JBIG. This speed up is comparable to the Group 4 standard, but with better compression performance. Merits Simplest This compression technique produces smoother, more satisfactory compressed images. Wavelet coding schemes at higher compression avoid blocking artifacts. They are better matched to the HVS (Human Visual System) characteristics. Compression with wavelets is scalable as the transform process can be applied to an image as many times as wanted and hence very high compression ratios can be achieved. Wavelet based compression allow parametric gain control for image softening and sharpening. Wavelet-based coding is more robust under transmission and decoding errors, and also facilitates progressive transmission of images. Wavelet compression is very efficient at low bit rates. Wavelets provide an efficient decomposition of signals prior to compression. The first rule guarantees that no more than the ceil(log2(alphabet size) rightmost bits of the code can differ from zero The first rule also allows an efficient decoding Both rules together allow a complete reconstruction of the code knowing only the code lengths for each symbol.

IMAGE COMPRESSION

3. SYSTEM DESIGN
The specification of a software product spell out what the product is to do. The aim of the design phase is to determine how to make the product when satisfies all the requirements given by requirement specification document. During the design phase, the internal structure of the product including the algorithm, data structures, inputs and outputs and interaction with external environment is finalized. So other words the entire software architecture of the produced is created. The design has to specify what each module has to perform which function and how to do it. Major design decisions are made in this phase. The decisions have to be documented in order to keep track of the entire design process. Unlike other engineering design approaches, software design changes conditionally as new techniques, new approaches evolve. In view of this software design requires a different treatment when compared to conventional engineering design methodologies. Further software design is an iterative process. 3.1 Input Design Input design is the most important part of the overall system design, which requires very careful attention, Often the collection of input data is the most expensive part of the system. Many errors occur during this phase of the design. So to make the system study, the inputs given by the user is strictly validated before making a manipulation with it thus, by validation it is possible to Provide an effective method of input

IMAGE COMPRESSION Achieve the highest possible level of accuracy Ensure that input is acceptable and understood by the user staff. Input design in mainly concentrated on estimating what the inputs are And how they have to be arranged on the input screen, how frequently the

10

Ideas are to be collected. The input screens are designed in such a manner that avoids confusion and guides the user in the correct way. Although study has been made on the type and how the input form is to be designed. Some inputs from the user may cause severe error and is strictly validated. This software provides a point to its user. A very good look and feel is provide through the organized arrangement of controls such as label box, textbox, Button, link button, object list, User Control etc. Input screen for Image Compression system are very simple and userfriendly User are allowed to access the software only after the user authentication process; if irrelevant data is entered message screens are displayed. a. Login Form This form is used to give the username and password. After that we enter into the next level. b. Method Form Using this form the user can choose which method is used to compress the image. c. Huffman Encoding Form This form is used to load the image and compress the image using Huffman algorithm and store the compressed image in desired location.

IMAGE COMPRESSION d. Huffman Decoding Form This form is used to decode the compressed image. e. Haar Discrete Wavelet Transform Form

11

This form is used to compress the image using Haar Discrete Wavelet Transform algorithm. 3.2 Output Design Output design generally refers to the results generated by the system. For many end-users, output is the main reason for developing the system and the basis on which they evaluate the usefulness of the application. The objective of a system finds its shape in terms of the output. The Analysis of the objective of a system leads to the determination of outputs.The Huffman Encoding form is used to view the original size of the image, size of the compressed image, Byte difference etc. The Haar Discrete Wavelet Algorithm form is used to view the size of the original image and compressed image. 3.3 Data Flow Diagram Context Level DFD User name USER Password LOGI N Huffman Method Loading Huffman Method Loading Harr Discrete Wavelet Haar Discrete Wavelet

IMAGE COMPRESSION DFD for Huffman Decode INPUT FILE Input image

12

DECODE Decoded image Decoded image

HAMMING

COMPRESSIO N Compressed image COMPRESSED IMAGE

DFD for Encode INPUT FILE

Input image ENCODE Encoded image ENCODED IMAGE

IMAGE COMPRESSION DFD for Method 2 INPUT FILE Input image

13

HARR DISCRETE

Compressed image COMPRESSED IMAGE

3.4 System Chart LOGIN

HUFFMAN

HAAR WAVELET

INPUT FILE

INPUT FILE DECODE COMPRESSION HAMMING ENCODE

COMPRESSION

IMAGE COMPRESSION

14

4. PROJECT DESCRIPTION
4.1 Detail Description Process Image compression is minimizing the size in bytes of a graphics file without degrading the quality of the image to an unacceptable level. The reduction in file size allows more images to be stored in a given amount of disk or memory space. It also reduces the time required for images to be sent over the Internet or downloaded from Web pages. In this project we use two methods of image compression. Huffman encoding. Haar Discrete Wavelet Transform algorithm these two methods are lossless compression technique. 4.2. Huffman Encoding In computer science and information theory, Huffman coding is an entropy encoding algorithm used for lossless data compression. The term refers to the use of a variable-length code table for encoding a source symbol (such as a character in a file) where the variable-length code table has been derived in a particular way based on the estimated probability of occurrence for each possible value of the source symbol. It was developed by David A. Huffman while he was a Ph.D. student at MIT, and published in the 1952 paper "A Method for the Construction of Minimum-Redundancy Codes".

IMAGE COMPRESSION

15

Huffman coding uses a specific method for choosing the representation for each symbol, resulting in a prefix code (sometimes called "prefix-free codes") (that is, the bit string representing some particular symbol is never a prefix of the bit string representing any other symbol) that expresses the most common characters using shorter strings of bits than are used for less common source symbols. Huffman was able to design the most efficient compression method of this type: no other mapping of individual source symbols to unique strings of bits will produce a smaller average output size when the actual symbol frequencies agree with those used to create the code. A method was later found to do this in linear time if input probabilities Refers to data compression techniques in which no data is lost. The PKZIP compression technology is an example of lossless compression. For most types of data, lossless compression techniques can reduce the space needed by only about 50%. For greater compression, one must use a lossy compression technique. Note, however, that only certain types of data graphics, audio, and video can tolerate lossy compression. You must use a lossless compression technique when compressing data and programs. Unlike ASCII code, which is a fixed-length code using seven bits per character, Huffman compression is a variable-length coding system that assigns smaller codes for more frequently used characters and larger codes for less frequently used characters in order to reduce the size of files being compressed and transferred. For example, in a file with the following data: XXXXXXYYYYZZ

IMAGE COMPRESSION

16

The frequency of "X" is 6, the frequency of "Y" is 4, and the frequency of "Z" is 2 shown in table 4.1.1. If each character is represented using a fixed-length code of two bits, then the number of bits required to store this file would be 24, i.e., (2 x 6) + (2x 4) + (2x 2) = 24. If the above data were compressed using Huffman compression, the more frequently occurring numbers would be represented by smaller bits, such as: Table 4.1.1.Character representation using fixed length X Y Z by the code 0 (1 bit) by the code 10 (2 bits) by the code 11 (2 bits)

therefore the size of the file becomes 18, i.e., (1x 6) + (2 x 4) + (2 x 2) = 18. In the above example, more frequently occurring characters are assigned smaller codes, resulting in a smaller number of bits in the final compressed file. The algorithm The Huffman coding is form of entropy coding which implements the coding idea

discussed above. If a random number has

possible outcomes (symbols), it is

obvious that these outcomes can be coded by

bits. For example, as the pixels in a

digital image can take

possible gray values, we need

to represent each

pixel and bits to represent an image of pixels.

IMAGE COMPRESSION By Huffman coding, however, it is possible to use on average fewer than

17 bits

to represent each pixel. In general, Huffman coding encodes a set of binary code of variable length, following this procedure:

symbols with

Estimate the probability

for each of the

symbols

Sort theses

in descending order (top down);

Forward pass (left to right):- combine the two smallest

at the bottom and resort their

sum with all other probabilities, repeat this step until there are only two probabilities left. Backward pass (right to left):- add a bit (0 or 1) to the binary codes of the two

probabilities newly emerging at each step, repeat this step until all the initial

symbols

are encoded. As the result of Huffman coding, all symbols are encoded optimally in the sense that more probable symbols are encoded by shorter binary codes, so that the average length (number of bits) of the codes for these symbols is minimized. To illustrate how much compression the Huffman coding can achieve and how the compression rate is related to the content of the image, consider the following examples

IMAGE COMPRESSION

18

of compressing an image of

gray levels, each with probability for a pixel

to be at the itch gray level

(the histogram of the image).

Example 4.1.1:

i.e., all pixels take the same value. The image contains 0 bits uncertainty (no surprise) or information and requires 0 bits to transmit. Example 4.1.2: Pi 0.5 0.5 0.0 0.0 The uncertainty is Code 0 1

which requires 1 bit to encode the information. Although the uncertainty is less than half of that in the previous case, it still requires one bit to encode the information.

IMAGE COMPRESSION

19

This image has the maximum information (uncertainty) and therefore requires bits to encode each of its pixels. From these examples it can be seen that with variable-length coding the average number of bits needed to encode a pixel may be reduced from. The amount of reduction, or the compression rate, depends on the amount of uncertainty or information contained in the image. If the image has only a single gray level, it contains 0 bits information and

requires 0 bits to encode; but if the image has maximum amount of Process bits information requiring

equally likely gray levels, it contains bits to encode.

Decoding: - The input is an array of bytes of the original file. The operations are: Calculating the entropy Calculating the entropy given another array of bytes (I called it twin but it should be a file from family) Decoding: decode the data using Algorithm Encoding: - The input is Decoded Info (data after compression). The operations are: Encoding data by reconstructing the original file Saving the result Loading compressed file Decoded Info: - This is the structure of the compressed file. It contains the following fields: decodedData: array of bytes containing the result of compression

IMAGE COMPRESSION codeMap: variable code which calculates using the Huffman's tree totalBits: size of the result

20

twinFileName: the name of the file which is used to calculate the xor result (it's the file that give the lowest entropy) originalFileName: just the name of the original file I use it when saving the result

Hamming:- The input is a histogram of the bytes in the original file. Using this information, we can build the trees which allow us to calculate the variable length code of the compressed data. This class is simply an implementation of the algorithm we learnt in class. 4.3. Haar Discrete Wavelet Transform algorithms Here we use the Haar Discrete Wavelet Transform. The Haar transform operates as a square matrix of length N = some integral power of 2. In order to implement the image compression algorithm we chose, we divided the process into various steps: calculate the sums and differences of every row of the image calculate the sums and differences of every column of the resulting matrix repeat this process until we get down to squares of 16x16 quantize the final matrix using different bit allocation schemes write the quantized matrix out to a binary file The next step is quantization. This is performed during the writing to a binary file (see proceeding discussion of writing to a file); however, we wrote a distinct quantization

IMAGE COMPRESSION

21

function to analyze this step statistically (MSE and PSNR -- see the Results section), as well as for our own educational benefit. The quantization function, quant, is also called in squishier. We used a number of different bit allocation masks in order to determine which scheme is better.

Wavelet-based compression Wavelet coding schemes at higher compression avoid blocking artifacts. They are better matched to the HVS (Human Visual System) characteristics. Compression with wavelets is scalable as the transform process can be applied to an image as many times as wanted and hence very high compression ratios can be achieved. Wavelet based compression allow parametric gain control for image softening and sharpening. Wavelet-based coding is more robust under transmission and decoding errors, and also facilitates progressive transmission of images. Wavelet compression is very efficient at low bit rates. Wavelets provide an efficient decomposition of signals prior to compression. Background Before we go into details of the method, we present some background topics of image compression which include the principles of image compression, the classification of compression methods and the framework of a general image coder and wavelets for image compression. Principles of Image Compression

IMAGE COMPRESSION

22

An ordinary characteristic of most images is that the neighboring pixels are correlated and therefore hold redundant information. The foremost task then is to find out less correlated representation of the image. Two elementary components of compression are redundancy and irrelevancy reduction.

Human Visual System (HVS) three types of redundancy can be identified: Spatial Redundancy or correlation between neighboring pixel values. Spectral Redundancy or correlation between different color planes or spectral bands. Temporal Redundancy or correlation between adjacent frames in a sequence of images especially in video applications. Image compression research aims at reducing the number of bits needed to represent an image by removing the spatial and spectral redundancies as much as possible.

Classification of Compression Technique There are two ways that we can consider for classifying compression techniques lossless vs. lossy compression Predictive vs. transform coding. Lossless vs. Lossy compression

IMAGE COMPRESSION

23

In lossless compression schemes, the reconstructed image, after compression, is numerically identical to the original image. However lossless compression can only achieve a modest amount of compression. An image reconstructed following lossy compression contains degradation relative to the original. Often this is because the compression scheme completely discards redundant information. However, lossy schemes are capable of achieving much higher compression. Under normal viewing conditions, no visible loss is perceived (visually lossless). Predictive vs. Transform coding In predictive coding, information already sent or available is used to predict future values, and the difference is coded. Since this is done in the image or spatial domain, it is relatively simple to implement and is readily adapted to local image characteristics. Differential Pulse Code Modulation (DPCM) is one particular example of predictive coding. Transform coding, on the other hand, first transforms the image from its spatial domain representation to a different type of representation using some well-known transform and then codes the transformed values (coefficients).This method provides greater data compression compared to predictive methods, although at the expense of greater computation. Framework of General Image Compression Method A typical lossy image compression system is consists of three closely connected components is shown in Figure 4.3.1. Source Encoder Quantizer Entropy Encode

IMAGE COMPRESSION INPUT IMAGE SOURCE ENCODER QUANTIZER ENTROPY ENCODER

24 COMPRESSED IMAGE

Figure4.3.1. Framework of General Image Compression Method Quantization can also be applied on a group of coefficients together known as Vector Quantization (VQ). Both uniform and non-uniform quantizes can be used depending on the problems. Wavelets for image compression Wavelet transform exploits both the spatial and frequency correlation of data by dilations (or contractions) and translations of mother wavelet on the input data. It supports the multiresolution analysis of data i.e. it can be applied to different scales according to the details required, which allows progressive transmission and zooming of the image without the need of extra storage. Another encouraging feature of wavelet transform is its symmetric nature that is both the forward and the inverse transform has the same complexity, building fast compression and decompression routines. Its characteristics well suited for image compression include the ability to take into account of Human Visual Systems (HVS) characteristics, very good energy compaction capabilities, robustness under transmission, high compression ratio etc. The implementation of wavelet compression scheme is very similar to that of sub band coding scheme: the signal is decomposed using filter banks. The output of the filter banks is down-sampled, quantized, and encoded. The decoder decodes the coded representation, up samples and recomposes the signal.

IMAGE COMPRESSION

25

Wavelet transform divides the information of an image into approximation and detail sub signals. The approximation sub signal shows the general trend of pixel values and other three detail sub signals show the vertical, horizontal and diagonal details or changes in the images. If these details are very small (threshold) then they can be set to zero without significantly changing the image. The greater the number of zeros the greater the compression ratio. If the energy retained (amount of information retained by an image after compression and decompression) is 100% then the compression is lossless as the image can be reconstructed exactly. This occurs when the threshold value is set to zero, meaning that the details have not been changed. If any value is changed then energy will be lost and thus lossy compression occurs. As more zeros are obtained, more energy is lost. Therefore, a balance between the two needs to be found out. Haar Wavelet Transform

To understand how wavelets work, let us start with a simple example. Assume we have a 1D image with a resolution of four pixels, having values [9 7 3 5]. Haar wavelet basis can be used to represent this image by computing a wavelet transform. To do this, first the average the pixels together, pair wise, is calculated to get the new lower resolution image with pixel values [8 4]. Clearly, some information is lost in this averaging process. We need to store some detail coefficients to recover the original four pixel values from the two averaged values. In our example, 1 is chosen for the first detail coefficient, since the average computed is 1 less than 9 and 1 more than 7. This single number is used to recover the first two pixels of our original four-pixel image. Similarly,

IMAGE COMPRESSION

26

the second detail coefficient is -1, since 4 + (-1) = 3 and 4 - (-1) =5. Thus, the original image is decomposed into a lower resolution (two-pixel) version and a pair of detail coefficients repeating this process recursively on the averages gives the full decomposition shown in Table 4.3.1. Table 4.3.1.Averages and detail coefficients

Resolution 4 2 1

Averages [9 7 3 5] [8 4] [6]

Detail Coefficients 1 -1 2

Thus, for the one-dimensional Haar basis, the wavelet transform of the original four-pixel image. We call the way used to compute the wavelet transform by recursively averaging and differencing coefficients, filter bank. Compression of 2D image with Haar Wavelet Technique It has been shown in previous section how one dimensional image can be treated as sequences of coefficients. Alternatively, we can think of images as piecewise constant functions on the half-open interval.. Let V0 be the vector space of all these functions. A two pixel image has two constant pieces over the interval. We call the space containing all these functions V1. If we continue in this manner, the space Vj will include all piecewise-constant functions defined on the interval with constant pieces over each of 2j equal subintervals. We can now think of every one-dimensional image with 2j pixels as an element, or vector, in Vj. Note that because these vectors are all functions defined on the unit interval, every vector in Vj is also contained in Vj+1. For example, we can always

IMAGE COMPRESSION

27

describe a piecewise constant function with two intervals as a piecewise-constant function with four intervals, with each interval in the first function corresponding to a pair of intervals in the second. It guarantees that every member of V0 can be represented exactly as a member of higher resolution space V1. The converse, however, is not true: not every function G(x) in V1 can be represented exactly in lower resolution space V0; in general there is some lost detail. Now we define a basis for each vector space Vj. The basic functions for the spaces V j is called scaling functions, and are usually denoted by the symbol . A simple basis for Vj is given by the set of scaled and translated box functions. ji(x) : = (2jx i) i = 0, 1, 2..2j -1 where

1 for 0x1 (x):= 0 otherwise The wavelets corresponding to the box basis are known as the Haar wavelets, given by ij (x) : = (2jx i) i = 0, 1, 2..2j -1 where

1 for 0x1/2 (x):= -1 for 1/2x<1 0 otherwise Thus, the DWT for an image as a 2D signal will be obtained from 1D DWT. We get the scaling function and wavelet function for 2D by multiplying two 1D functions. The scaling function is obtained by multiplying two 1D scaling functions:

IMAGE COMPRESSION

28

(x,y)=(x)(y). The wavelet functions are obtained by multiplying two wavelet functions or wavelet and scaling function for 1D. Table 4.3.2.Structure of wavelet decomposition

LL LH3 LH2

HL3 HH3 HL2 HH2 LH1 HH1 HL1

It applies the 1D wavelet transform to each row of pixel values shown in Figure 4.3.3. This operation provides us an average value along with detail coefficients for each row. Next, these transformed rows are treated as if they were themselves an image and apply the 1D transform to each column. Now let us see how the 2D Haar wavelet transformation is performed shown in Figure 4.3.4. The image is comprised of pixels represented by numbers. Consider the 88 image taken from a specific portion of a typical image as shown in Figure 4.3.5. Let us look how the operation is done.

Figure 4.3.3 1-Level Of Decomposition

IMAGE COMPRESSION

29

Figure 4.3.4 2-Level Of Decomposition

Averaging: (64+2)/2=33, (3+61)/2=32, (60+6)/2=33, (7+57)/2=32 Differencing: 6433 =31, 332= 29, 6033=27 and 732= 25

Figure 4.3.3.A 88 image 2D Array Representation 64 9 17 40 32 41 49 8 2 55 47 26 34 23 15 58 3 54 46 27 35 22 14 59 61 12 20 37 29 44 52 5 60 13 21 36 28 45 53 4 6 51 43 30 38 19 11 62 7 50 42 31 39 18 10 63 57 16 24 33 25 48 56 1

IMAGE COMPRESSION

30

So, the transformed row becomes (33 32 33 32 31 29 2725). Now the same operation on the average values i.e. (33 32 33 32) is performed. Then we perform the same operation on the averages i.e. first two elements of the new transformed 31 29 27 25). The new matrix we get after applying this operation on each row of the entire matrix performing the same operation on each column of the matrix we get the final transformed matrix. This operation on rows followed by columns of the matrix is performed recursively depending on the level of transformation meaning the more iteration provides more transformations. Note that the left-top element of the i.e. 32.5 is the only averaging element which is the overall average of all elements of the original matrix and the rest all elements are the details coefficients.

The main part of the C program used to transform the matrix is shown in. The 2D array mat holds the values which represent the image. The point of the wavelet transform is that regions of little variation in the original image manifest themselves as small or zero elements in the wavelet transformed version. A matrix with a high proportion of zero entries is said to be sparse. For most of the image matrices, their corresponding wavelet transformed versions are much sparser than the originals. Very sparse matrices are easier to store and transmit than ordinary matrices of the same size. This is because the sparse matrices can be specified in the data file solely in terms of locations and values of their non-zero entries. It can be seen that in the final transformed matrix, we find a lot of entries zero. From this transformed matrix, the original matrix can be easily calculated just by the reverse operation of averaging and differencing i.e. the original image can be reconstructed from the transformed image without the loss of information.

IMAGE COMPRESSION

31

Thus, it yields a lossless compression of the image. However, to achieve more degree of compression, we have to think of the lossy compression. In this case, a nonnegative threshold value say e is set.

. These thresholding methods are defined as follows: 0, if | x | < e .. (Hard Thresholding) T(e,x) = x, otherwise

0,if | x |< e . (Soft Thresholding T(e, x) = Sign(x)(| x |-e otherwise

0, if x< (2log2N) (Universal Thresholding) T(e,x)= x, otherwise

Where is the standard deviation of the wavelet coefficients and N is the number of wavelet coefficient. In summary, the main steps of the 2D image compression using Haar Wavelet as the basis functions are
Start with the matrix P representing the original image,

IMAGE COMPRESSION

32

Compute the transformed matrix T by the operation averaging and differencing

(First for each row, then for each column)

Choose a threshold method and apply that to find the new matrix say D Use D to compute the compression ratio and others values and to reconstruct the

original image as well. A database of twenty gray scale images each with size 256256 is used in the experiment. We define the compression ratio (CR) as the ratio of the number of nonzero elements in original matrix to the number of nonzero elements in updated transformed matrix. The enthusiastic CR values for different thresholding methods and different is shown in Table 4.3.3. Table 4.3.3 Thresholding methods

e
15 20 25

CR(Hard Threshold) 15.74 17.11 18.47

CR(Soft Threshold) 14.1 15.87 16.95

The universal thresholding method generates the CR as 14.375. It is noted here that the hard thresholding provides the best CR. The soft thresholding gives better CR in comparison to universal thresholding method but it depends on choosing the value of .

IMAGE COMPRESSION Code For Transformation /*row transformation*/ for(i=0;i<row;i++){w=col; do{ k=0; /*averaging*/ for(j=0;j<w/2;j++) a[j]=((mat[i][j+j]+mat[i][j+j+1])/2); /*differencing*/ for(j=w/2;j<w;j++,k++) a[j]=mat[i][j-w/2+k]-a[k]; for(j=0;j<row;j++) mat[i][j]=a[j]; w=w/2; }while(w!=1); } /*column transformation*/ for(i=0;i<col;i++){ w=row; do{k=0; /*averaging*/ for(j=0;j<w/2;j++) a[j]=((mat[j+j][i]+mat[j+j+1][i])/2); /*differencing*/ for(j=w/2;j<w;j++,k++) a[j]=mat[j-w/2+k][i]-a[k]; for(j=0;j<w;j++) mat[j][i]=a[j]; w=w/2; }while(w!=1)

33

IMAGE COMPRESSION The PSNR for gray scale image (8 bits/pixel) is defined by-

34

Where MSE is the Man Squared Error defined by-

where I is original image, I1 is approximation of decompressed image and m, nare dimensions of the image. The PSNR values for different threshold values and techniques are. The soft thresholding method performs better than hard thresholding shown in Figure 4.3.6. The universal method reports PSNR as 24.875. These results are very much acceptable in most cases except in medical application where no loss of information is to be guaranteed. However, the PSNR is not adequate as a perceptually meaningful measure of picture quality, because the reconstruction errors generally do not have the characteristic of signal independent additive noise and the seriousness of the impairments cannot be measured by a simple power measurement. Small impairment of an image can lead to a very small PSNR in lieu of the fact that the perceived image quality can be acceptable. So, the perceptual quality measurement method quantified by MOS and PQS has been applied. The reference and test conditions are arranged in pairs such that the first is the unimpaired reference and the second is the same sequence impaired.

IMAGE COMPRESSION

35

Figure 4.3.6.Soft and Hard Thresholding The original image without compression Hard Threshold oppression was used as the reference condition. The viewers are asked to vote on the second, keeping in mind the first. The method uses the five grade impairment scale: 5 (Excellent), 4 (Good), 3 (Slightly annoying), 2 (Annoying) and 1 (Very annoying). At the end, the MOS is calculated as-

Where i is grade and p(i) is grade probability. PQS defined by-

IMAGE COMPRESSION

36

Uses some properties of HVS relevant to global image impairments such as random errors and emphasizes the perceptual importance of structured and localized errors. Here, a linear combination of uncorrelated principal distortion measures Zi combined by partial regression coefficients bi are used. PQS is constructed by regressions with MOS. The MOS and PQS values obtained are tabulated below which are very much encouraging. The number of decompositions determines the quality of compressed image. The number of decompositions also determines the resolution of the lowest level in wavelet domain. If a larger number of decompositions are used, it will provide more success in resolving important DWT coefficients from less important coefficients. The HVS is less sensitive to removal of smaller details. Table 4.3.4 Quality measurement methods.

e=15 MOS PQS

Hard Thresholding Soft Thresholding Universal Thresholding

MOS 4.675 4.80 4.865

PQS 4.765 4.875 4.957

After decomposing the image and representing it with wavelet coefficients, compression can be performed by ignoring all coefficients below some threshold.

Compression algorithm provides two modes of operation:

IMAGE COMPRESSION

37

Compression ratio is fixed to the required level and threshold value has been changed to achieve required compression ratio; after that, PSNR is computed.
PSNR is PSNR is fixed to the required level and threshold values have been

changed to achieve required PSNR; after that, CR is computed. It is noted that image quality is better for a larger number of decompositions. On the contrary, a larger number of decompositions causes the loss of the coding algorithm efficiency. Therefore, adaptive decomposition is required to achieve balance between image quality and computational complexity. PSNR tends to saturate for a larger number of decompositions. For each compression ratio, the PSNR characteristic has threshold which represents the optimal number of decompositions At present, the most widely used objective distortion measures are the MSE and the related PSNR. They can easily be computed to represent the deviation of the distorted image from the original image in the pixel wise sense. However, in practical viewing situations, human beings are usually not concentrated on pixel differences alone, except for particular applications such as medical imaging, where pixel wise precision can be very important. The subjective perceptual quality includes surface smoothness, edge sharpness and continuity, proper background noise level, and so on. Image compression techniques induce various types of visual artifacts that affect the human viewing experience in many distinct ways, even if the MSE or PSNR level is adjusted to be about equal.

It is generally agreed that MSE or PSNR does not correlate well with the visual quality perceived by human beings, since MSE is computed by adding the squared differences of individual pixels without considering the spatial interaction among

IMAGE COMPRESSION

38

adjacent pixels. Some work tries to modify existing quantitative measures to accommodate the factor of human visual perception. One approach is to improve MSE by putting different weights to neighboring regions with different distances to the focal pixel. Most approaches can be viewed as curve-fitting methods to comply with the rating scale method. In order to obtain an objective measure for perceived image fidelity, models of the human visual system (HVS) should be taken into account. It is well known that the HVS has different sensitivities to signals of different frequencies. Since the detection mechanisms of the HVS have not localized responses in both the space and frequency domains, the space-based MSE nor the global Fourier analysis provides a good tool for the modeling. So, here the perceptual quality measurement method quantified by MOS and PQS has been applied and the results are encouraging shown in Figure 4.3.4. The fundamental difficulty in testing an image compression system is how to decide which test images to use for evaluation. The image content being viewed influences the perception of quality irrespective of technical parameters of the compression system. A series of pictures which are average in terms of how difficult they are for system being evaluated has been selected. In this paper, only the gray-scale images are considered. However, wavelet transforms and compression techniques are equally applicable to color images with three color components. We have to perform the wavelet transform independently on each of the three color components of the images and have to treat the results as an array of vectored valued wavelet co-efficient. In this case, in lieu of using the absolute value of the scalar

IMAGE COMPRESSION

39

co-efficient, a vector-valued coefficient is to be used. Furthermore, a number of ways can be used in which the color information can be used to obtain a wavelet transform that is even sparser For example, by first converting the pixel values in an image from RGB colors to YIQ colors, we can separate the luminance information (Y) from chromatic information (I and Q). Once the wavelet transform is computed, the compression method can be applied to each of the components of the image separately. Since the human perception is most sensitive to variation in Y and least sensitive in Q, the compression scheme may be permitted to tolerate a larger error in the Q component of the compressed image, thereby increasing the scale of compression.

5. IMPLEMENTATION
5.1 About the Software

IMAGE COMPRESSION

40

The .NET Framework is a new computing platform that simplifies application development in the highly distributed environment of the internet. The .NET Framework is designed to fulfill the following objectives: To provide a consistent object-oriented programming environment whether the object code is stored and executed locally, executed locally but Internetdistributed, or executed remotely. To provide a code-execution environment that minimizes the software development and versioning conflicts. To make the developer experience consistent across widely varying types of applications, such as Windows-based applications and Web-based applications. The .NET Framework has two main components: The common language runtime The .NET Frame work class library .NET Framework is a component of the Microsoft Windows Operating System used to build and run windows based applications. One feature of the .NET Framework that saves us from huge amount of tedious coding is the base class library. The base framework classes cover a multitude of different functions. These functions are all provided in a number of base classes that are grouped together under a namespace. The .NET Framework is the infrastructure for the Microsoft .NET platform. The .NET Framework is an environment for building, deploying, and running Web applications and Web Services. The Microsoft .NET Framework was developed to solve this problem.

IMAGE COMPRESSION .NET Frameworks keywords: Easier and quicker programming Reduced amount of code Declarative programming model Richer server control hierarchy with events Larger class library Better support for development tools The .NET Framework consists of 3 main parts: Programming languages: C# (pronounced C sharp) Visual Basic (VB.NET) J# (pronounced j sharp) Server technologies and client technologies: ASP.NET (Active Server Page) Windows Forms (Windows Desktop Solutions) Compact Framework (PDA/Mobile solutions Development environments: Visual Studio.NET (VS.NET) Visual Web Developer Intermediate Language (MSIL)

41

The programmer can use any .NET language to write the code including Visual Basic (VB), c#, j script etc. The result is then compiled to MSIL, the common language of .NET.

IMAGE COMPRESSION Common Language Runtime (CLR)

42

The CLR is described as the execution engine of .Net. Its this CLR that manages the execution of programs. It provides the environment within which the programs run. The software version of .Net is actually the CLR version. Working of the CLR When the .Net program is compiled, the output of the compiler is not an executable file but a file that contains a special type of code called the Microsoft Intermediate Language (MSIL). This MSIL defines a set of portable instructions that are independent of any specific CPU. Its the job of the CLR to translate this Intermediate code into an executable code when the program is executed making the program to run in any environment for which the CLR is implemented and thats how the .Net Framework achieves portability. This MSIL is turned into executable code using a JIT (Just In Time) compiler. Class Libraries Class library is the second major entity of the .Net Framework. This library gives the program access to runtime environment. The class library consists of lots of prewritten code that all the applications created in VB.net and Visual Studio.Net will use. Common Language specification (CLS) If we want the code which we write in a language to be used by programs in other languages then it should adhere to the Common Language Specification (CLS). The CLS describes a set of features that different languages have in common.

IMAGE COMPRESSION Some reasons why developers are building applications using the .NET Framework: Improved Reliability Increased Performance Developer Productivity Powerful Security Integration with existing Systems Ease of Deployment Mobility Support XML Web service Support Support for over 20 Programming Languages Flexible Data Access Languages supported by .NET Framework The table below lists all the languages supported by .NET Framework and describes those languages: APL

43

APL is one of the most powerful, consistent and concise programming languages ever devised. It is a language for describing procedures in the processing of information. It can be used to describe mathematical procedures having nothing to do with computers or to describe the way a computer works. C++ C++ is a true OOP. It is one of the early Object-Oriented programming Languages. C++ derives from the C language.

IMAGE COMPRESSION VC++

44

Visual C++ is the name of a C++ compiler with an integrated environment from Microsoft. This includes special tools that simplify the development of great applications, as well as specific libraries. Its use is known as Visual programming. C# C# called as C Sharp is a fully fledged object-Oriented programming language from Microsoft built into the .Net Framework. First created in the late 1990s was part of Microsofts whole .Net strategy. COBOL COBOL (Common Business Oriented Language) was the first widely-used highlevel programming language for business applications. It is considered as a programming language to have more lines of code than any other language.

IMAGE COMPRESSION

45

6. TESTING
6.1 Testing Fundamentals There are two types of testing; the first one is unit testing where the important classes are tested by writing a test application. Here all the methods of the classes are tested for its defined functionality. The next level testing is called black box testing or integration testing, where the software is tested as whole for its functionalities. This kind of doesnt need any inside knowledge of the system but the tester should know about how to work with the software and the functionalities provided. Unit test focuses verification effort on the simplest unit of software design of the module using the detailed writing description as a guide, internal control path are tested to uncover errors within the boundary of module. The unit test is always white box oriented and the step can be conducted in parallel for multiple modules. 6.2. Type of Testing Done Testing forms a core part of any project there are various types of testing are there. In this system we are done following testing White Box testing Black box testing Unit Testing Integration Testing Validation Testing Output Testing User Interface Testing Data Testing

IMAGE COMPRESSION Execution Testing 6.2.1. White Box Texting

46

White-box testing, sometimes called glass-box is a test case design method that uses the control structure of the procedural design to derive test cases. 6.2.2. Block Box Testing Focuses on the functional requirements of the software That is, black box enables the software engineer to derive sets of input conditions that will fully exercise all functional requirements for the program. Black box testing is not an alternative to white box testing. Rather it is a complementary approach that is likely to uncover a different class of Errors than white box methods. Black Box testing attempts to find errors in the following categories: Incorrect or missing functions Interface errors 6.2.3. Unit Testing This is the first level of testing. In this different modules are tested against the specifications produced during the design of the modules. Unit testing is done for the verification of the code produced during the coding of single program modules is an isolated environment. Unit testing first focuses on the modules independently of one another to locate errors. After coding each dialog is tested and run individually. All necessary coding was removed and ensured that all the modules worked, as the programmer would expect. Logical errors found were corrected.

IMAGE COMPRESSION

47

6.2.4. Integration Testing Data can be lost access an interface, one module can have as adverse effort on the other sub functions when combined may not produce the desired major functions. Integration testing is a systematic testing for constructing the program structure, while at the same time conducting tests to uncover error associated with the interface. The objectives are to take a unit testing as a whole. Here the correction is difficult because vast expenses of the entire program complicate the isolation of causes. Thus in the integration testing step, all the errors are uncovered are corrected for the next testing steps. 6.2.5. Validation Testing This provides the final assurance that the software meets all functional, behavioral and performances requirements. The software is completely assembled as a package. Validation succeeds when the software functions in a manner in which the user expects. Validation refers to the process of using software in alive environment in order to find errors. During the course of validating the system, failures may occur and sometimes the coding has to be changed according to the requirement. Thus the feedback from the validation phase generally produces changes in the software. Once the application was made free all of logical and interface errors, inputting dummy data ensured that the software developed satisfied all the requirements of the user.

IMAGE COMPRESSION

48

6.2.6. Output Testing After performing the Validation testing, the next step is output testing of the proposed systems since no system could be useful if it does not produces the required output generated or considered into two ways; one is no screen and another is printed format. The output format on the screen is found to be correct as the format was designed in the system design phase according to the user needs. For the hard copy also the output comes out as the specified requirements by the user. Hence output testing does not result in any correction in the system. 6.2.7. User Interface Testing User Interface of a system is the key factor for the success of any system. The system under consideration is tested for user acceptance by constantly keeping in touch with the prospective system users at the time of developing and making changes whenever required. Preparation of test data plays a vital role in the system testing. After preparing the test data the system under study is tested using the test data. While testing the system by using test data errors are again uncovered and the corrections are also noted for future use. 6.2.8. Data Testing After preparing the test data the system under study was tested using test data. While testing the system by using test data, errors were again uncovered and corrected by using the above testing steps. Preparation of test data plays a vital role in the system testing. Taking various types of test data does all the above testing.

IMAGE COMPRESSION

49

6.2.9. Execution Testing Test data was prepared which were the acknowledgement details and the information regarding the various departments in the case. An already existing file was taken from the database and the data was fed into the new system. Various tests as mentioned above were carried out. Initially there were bug drawbacks for the user to complete the same process. Those bugs and drawbacks were noted down and modified later. Again the same process was repeated three to four times. All the out puts generated were compared with the existing file documents and the newly developed package is running properly and efficiently.

IMAGE COMPRESSION

50

7. USER MANUAL
7.1 Hardware Specification The hardware for the system is selected considering the factors such as CUP processing speed, memory access speed, peripheral channel speed, seek time & relational delay of hard disk and communication speed etc. Processor RAM HDD Display Keyboard Mouse 7.2 Software Specification Operation System Front-end Code-behind -Windows XP - ASP.Net - VB.Net, C# - Pentium IV - 512 MB - 8O GB - Color Monitor

IMAGE COMPRESSION

51

8. FUTURE ENHANCEMENTS
In our project we are doesnt use database, so we cant perform compression over network. Here in this project the enhancements include embedding of the security

features into this system. Now in this stage of project the security issues are addressed by implementing the session features in the ASP.NET, but in this world the security issues have to be addressed in the better and secure manner, like embedding the hashing algorithms into the solution. In our Huffman compression technique we cant able to show the preview of selected image, and Harr Discrete method is only perform the compression that does not perform reconstruct compressed image into its original image. According to the needs arising in the long run time many additional options can be added to the A STUDY ON LOSSLESS IMAGE COMPRESSION solution. This solution can also be enhanced by adding new features with the existing ones.

IMAGE COMPRESSION

52

9. CONCLUSION
This system has been developed for the given conditions and is found working effectively. The developed system is flexible and changes, whenever needed, can be made easily. The software is developed with a very good user interface and in a simple manner using VB.net so that even the first time user can user can use it without any problem. The system is highly scalable and is well efficient to make easy interactions with the client. The pages opened on the solution are small in size so that the end user doesnt have to worry about the download time and also they are fast in getting a response. The user-friendly interface successfully overcomes strict and severe validation checks, using the test data. The results attained were fully satisfactory from the user point of view. An attempt was made to obtain maximum perfection in documenting the software in a simple, precise and self-explanatory manner. The system was verified with valid as well as invalid data in each manner. The system is done with an insight into the necessary modifications that may require in the future. Hence the system can be maintained successfully without much rework.

IMAGE COMPRESSION

53

APPENDIX A
Screen Layouts Login Form

Enter User Name And Password

IMAGE COMPRESSION

54

Methods Form

Select Huffman

IMAGE COMPRESSION

55

Huffman Compression Home Page

IMAGE COMPRESSION

56

Adding Files

IMAGE COMPRESSION

57

Decoding

Saving Outputs

IMAGE COMPRESSION After Hamming

58

After Huffman

Encode

IMAGE COMPRESSION

59

Adding File

After Adding File

IMAGE COMPRESSION

60

Saving Encoded File

Harr Discrete Method

IMAGE COMPRESSION Selecting Harr Discrete Method

61

Harr Discrete Home Page

Adding File

IMAGE COMPRESSION

62

After Adding File

Compressing And Saving

IMAGE COMPRESSION

63

After Compressing

APPENDIX B

IMAGE COMPRESSION Source Code Login using System.Drawing; using System.Text; using System.Windows.Forms; namespace FamilyCompression { public partial class Login : Form { public Login() { InitializeComponent(); } private void cmdlogin_Click(object sender, EventArgs e) { if (txtuname.Text == "admin" && txtpass.Text == "welcome") { frmmethod frm = new frmmethod(); frm.Show(); // this.Close(); } } private string ucase(string p) { throw new Exception("The method or operation is not implemented."); }}}

64

Select Method 1 Huffman

IMAGE COMPRESSION
using System; using System.Collections.Generic; using System.ComponentModel; using System.Data; using System.Drawing; using System.Text; using System.Windows.Forms; namespace FamilyCompression { public partial class frmmethod : Form { public frmmethod() { InitializeComponent(); } private void cmdlogin_Click(object sender, EventArgs e) { if (optmethod1.Checked == true) { FamilyCompression family = new FamilyCompression(); family.Show(); this.Close(); } else {

65

System.Diagnostics.Process.Start(@"C:\Users\VIJIL\Desktop\Image\ImageCompression\ImageC ompression\obj\Debug\ImageCompression.exe"); } } private void optmethod1_CheckedChanged(object sender, EventArgs e) { } } }

Family Compression

IMAGE COMPRESSION
using System; using System.Collections.Generic; using System.ComponentModel; using System.Data; using System.Drawing; using System.Text; using System.Windows.Forms; using System.IO; namespace FamilyCompression { public partial class FamilyCompression : Form { #region Member Variables private DecodedInfo hammingRes = null; private DecodedInfo familyRes = null; private EncodedInfo encodedInfo = null; #endregion // Member Variables #region Construction public FamilyCompression() { InitializeComponent(); this.SetControlsState(false); } // FamilyCompression #endregion // Construction #region Private Methods private bool ReadInputFile(out byte[] inputData, string fileName) { inputData = null; FileStream inputFile = null; try { inputFile = new FileStream(fileName, FileMode.Open); if (inputFile.Length == 0) {

66

IMAGE COMPRESSION

67

MessageBox.Show("Please don't load an empty file!", "Invalid File", MessageBoxButtons.OK, MessageBoxIcon.Error); inputFile.Close(); return false; } inputData = new byte[inputFile.Length]; inputFile.Read(inputData, 0, (int)inputFile.Length); inputFile.Close(); } catch (Exception e) { MessageBox.Show("Failed to open the input file.\n\n\n" + e, "Opening file", MessageBoxButtons.OK, MessageBoxIcon.Error); this.SetControlsState(false); try { inputFile.Close(); } catch { } return false; } return true; } // ReadInputFile private void SaveOutputFile(byte[] data, string defultFileName) { if (data == null) { return; } SaveFileDialog dlg = new SaveFileDialog(); dlg.AddExtension = true; dlg.Filter = "All files (*.*)|*.*"; dlg.Title = "Save Data After Hamming"; dlg.FileName = defultFileName; if (dlg.ShowDialog() == DialogResult.OK) { try

IMAGE COMPRESSION
{ FileStream fs = new FileStream(dlg.FileName, FileMode.CreateNew); fs.Write(data, 0, data.Length); fs.Close(); } catch (Exception e) {

68

MessageBox.Show("Failed to save the file.\n\n\n" + e, "Saving file", MessageBoxButtons.OK, MessageBoxIcon.Error); } } } // SaveOutputFile private void SetControlsState(bool state) { if (!state) { this.ResetControls(); } this.buttonDecode.Enabled = state; this.buttonEncode.Enabled = state; this.groupBoxDetails.Enabled = state; this.labelProcedure.Enabled = state; this.groupBoxSaveResults.Enabled = state; this.progressBarProcedure.Enabled = state; this.buttonSaveOriginalData.Enabled = false; this.labelInputFileEntropy.Visible = false; this.labelInputFileEntropy.ForeColor = Color.Black; this.Refresh(); } // SetControlsState private void ResetControls() { this.progressBarOriginalFileSize.Value = 0; this.progressBarAfterHuffmanSize.Value = 0; this.progressBarTheNewMethodSize.Value = 0; this.labelOriginalFileSize.Text = "0"; this.labelAfterHuffmanSize.Text = "0";

IMAGE COMPRESSION
this.labelTheNewMethodSize.Text = "0"; this.textBoxInputFile.Text = string.Empty; this.buttonBrowse.Select(); } // ResetControls private string[] OpenFile(bool multiselect) { OpenFileDialog dlgOpenFile = new OpenFileDialog(); dlgOpenFile.AddExtension = true; dlgOpenFile.CheckFileExists = true; dlgOpenFile.DefaultExt = ""; string filter = "All files (*.*)|*.*"; dlgOpenFile.Filter = filter; dlgOpenFile.Title = "Open File"; dlgOpenFile.Multiselect = multiselect; if (dlgOpenFile.ShowDialog() == DialogResult.OK) { return dlgOpenFile.FileNames; } return null; } // OpenFile #endregion // Private Methods #region Public Methods #endregion // Public Methods #region Events

69

Adding Files Browse button

private void buttonBrowse_Click(object sender, EventArgs e) { string[] fn = this.OpenFile(false); if (fn != null) { this.SetControlsState(true); this.textBoxInputFile.Text = fn[0]; }

IMAGE COMPRESSION
} // buttonBrowse_Click

70

Add File
private void buttonAddFile_Click(object sender, EventArgs e) { string[] fn = this.OpenFile(false); string im=""; if (fn != null) { this.SetControlsState(true); for (int i = 0; i < fn.Length; i++) { ListViewItem itm = this.listViewFilesFamily.Items.Add(fn[i]); textBoxInputFile.Text = fn[i]; im = textBoxInputFile.Text; } }

Decode
private void buttonDecode_Click(object sender, EventArgs e) { this.progressBarProcedure.Maximum = this.listViewFilesFamily.Items.Count; byte[] data = null; if (!this.ReadInputFile(out data, this.textBoxInputFile.Text)) { return;} this.progressBarOriginalFileSize.Maximum = data.Length; this.progressBarOriginalFileSize.Value = data.Length; this.labelOriginalFileSize.Text = this.progressBarOriginalFileSize.Value.ToString(); this.labelInputFileEntropy.ForeColor = Color.Black; this.labelProcedure.Text = "Calculating Intropy.."; this.Refresh(); for (int i = 0; i < this.listViewFilesFamily.Items.Count; i++)

IMAGE COMPRESSION
{ try { this.listViewFilesFamily.Items[i].SubItems[1].Text = string.Empty; } catch { } } Decoding decodingHamming = new Decoding(data); double H = decodingHamming.CalcEntropy(); double[] Hs = new double[this.listViewFilesFamily.Items.Count]; this.labelInputFileEntropy.Text = H.ToString(); this.labelInputFileEntropy.Visible = true; this.Refresh(); string twinFileName = string.Empty; for (int i = 0; i < this.listViewFilesFamily.Items.Count; i++) { twinFileName = this.listViewFilesFamily.Columns[0].ListView.Items[i].Text; if (!this.ReadInputFile(out data, twinFileName)) { break; } Hs[i] = decodingHamming.CalcEntropyByTwin(data); try { this.listViewFilesFamily.Items[i].SubItems[1].Text = Hs[i].ToString(); } catch { this.listViewFilesFamily.Items[i].SubItems.Add(Hs[i].ToString()); } this.progressBarProcedure.Value = i + 1; this.Refresh(); } int minIdx = 0; for (int i = 1; i < Hs.Length; i++) {

71

IMAGE COMPRESSION
if (Hs[minIdx] > Hs[i]) { minIdx = i; } } try { twinFileName = this.listViewFilesFamily.Columns[0].ListView.Items[minIdx].Text; } catch { } if (Hs.Length > 0 && Hs[minIdx] < H) { this.listViewFilesFamily.Items[minIdx].BackColor = Color.Pink; if (!this.ReadInputFile(out data, twinFileName)) { this.ResetControls(); return; } } else { this.labelInputFileEntropy.ForeColor = Color.Red; data = null; } this.labelProcedure.Text = "Decoding.."; this.progressBarProcedure.Value = 0; this.Refresh(); this.hammingRes = new DecodedInfo(); decodingHamming.Decode(decodingHamming.InputData, out this.hammingRes, this.progressBarProcedure); string[] tempSplit = this.textBoxInputFile.Text.Split('\\'); this.hammingRes.originalFileName = tempSplit[tempSplit.Length - 1]; if (data != null) { this.familyRes = new DecodedInfo(); byte[] twin = (byte[])decodingHamming.InputData.Clone(); Compression.Xor(twin, data, twin);

72

IMAGE COMPRESSION
Decoding decodingTwin = new Decoding(twin); decodingTwin.Decode(twin, out this.familyRes, this.progressBarProcedure); this.familyRes.twinFileName = twinFileName; this.familyRes.originalFileName = tempSplit[tempSplit.Length - 1]; } else { this.familyRes = null; } this.labelAfterHuffmanSize.Text = this.hammingRes.decodedData.Length.ToString(); this.progressBarAfterHuffmanSize.Maximum = this.progressBarOriginalFileSize.Maximum; this.progressBarAfterHuffmanSize.Value = Math.Min(this.hammingRes.decodedData.Length, this.progressBarAfterHuffmanSize.Maximum - 1); this.buttonSaveDataAfterHamming.Enabled = true; if (this.familyRes != null) { this.labelTheNewMethodSize.Text = this.familyRes.decodedData.Length.ToString();

73

this.progressBarTheNewMethodSize.Maximum = this.progressBarOriginalFileSize.Maximum; this.progressBarTheNewMethodSize.Value = Math.Min(this.familyRes.decodedData.Length, this.progressBarTheNewMethodSize.Maximum - 1); this.buttonSaveDataNewMethod.Enabled = true; } else { this.labelTheNewMethodSize.Text = "0"; this.progressBarTheNewMethodSize.Value = 0; this.buttonSaveDataNewMethod.Enabled = false; } this.progressBarProcedure.Value = 0; this.labelProcedure.Text = "Ready"; double x1 = (double)Convert.ToDouble(this.labelOriginalFileSize.Text); double x2 = (double)Convert.ToDouble(this.labelAfterHuffmanSize.Text); double x3 = (double)Convert.ToDouble(this.labelTheNewMethodSize.Text); double y1 = x1 - x2; double y2 = x1 - x3; this.textBox1.Text = y1.ToString();

IMAGE COMPRESSION
this.textBox2.Text = y2.ToString(); } // buttonDecode_Click

74

Save After Hamming

private void buttonSaveDataAfterHamming_Click(object sender, EventArgs e) { SaveFileDialog dlg = new SaveFileDialog(); dlg.AddExtension = true; dlg.Filter = "Compressed files (*.ham)|*.ham|All files (*.*)|*.*"; dlg.Title = "Save Data After Hamming"; if (dlg.ShowDialog() == DialogResult.OK) { try { Encoding.Save(this.hammingRes, dlg.FileName); } catch (Exception ex) { MessageBox.Show("Failed to save the file.\n\n\n" + ex, "Saving file", MessageBoxButtons.OK, MessageBoxIcon.Error); } }} // buttonSaveDataAfterHamming_Click

Save After Huffman

private void buttonSaveDataNewMethod_Click(object sender, EventArgs e) { SaveFileDialog dlg = new SaveFileDialog(); dlg.AddExtension = true; dlg.Filter = "Compressed files (*.ham)|*.ham|All files (*.*)|*.*"; dlg.Title = "Save Data After the new method"; if (dlg.ShowDialog() == DialogResult.OK) { try {

IMAGE COMPRESSION
Encoding.Save(this.familyRes, dlg.FileName); } catch (Exception ex) { MessageBox.Show("Failed to save the file.\n\n\n" + ex, "Saving file", MessageBoxButtons.OK, MessageBoxIcon.Error); } } } // buttonSaveDataNewMethod_Click

75

Clear
private void buttonClear_Click(object sender, EventArgs e) { this.listViewFilesFamily.Items.Clear(); this.textBoxInputFile.Clear(); this.progressBarOriginalFileSize.Value = 0; this.progressBarAfterHuffmanSize.Value = 0; this.progressBarTheNewMethodSize.Value = 0; this.labelOriginalFileSize.Text = ""; this.labelAfterHuffmanSize.Text = ""; this.labelTheNewMethodSize.Text = ""; this.textBox1.Clear(); this.textBox2.Clear(); labelInputFileEntropy.Text=""; } // buttonClear_Click

Encode
private void buttonEncode_Click(object sender, EventArgs e) { try { DecodedInfo inputDecodedInfo = null; Encoding.Load(out inputDecodedInfo, this.textBoxInputFile.Text); this.labelProcedure.Text = "Encoding.."; EncodedInfo enc = new EncodedInfo();

IMAGE COMPRESSION
enc.ecodedData = Encoding.Encode(inputDecodedInfo, this.progressBarProcedure); enc.originalFileName = inputDecodedInfo.originalFileName; this.buttonSaveOriginalData.Enabled = inputDecodedInfo != null; this.encodedInfo = enc; } catch (Exception ex) { MessageBox.Show("Failed to encode the input file.\n\n\n" + ex, "Encoding", MessageBoxButtons.OK, MessageBoxIcon.Error); }this.labelProcedure.Text = "Ready.."; this.progressBarProcedure.Value = 0; } // buttonEncode_Click

76

Saving After Encode

private void buttonSaveOriginalData_Click(object sender, EventArgs e) { try { if (this.encodedInfo == null) { throw new Exception("Bad file format."); } this.SaveOutputFile(this.encodedInfo.ecodedData, this.encodedInfo.originalFileName); } catch (Exception ex) { MessageBox.Show("Failed to save the file.\n\n\n" + ex, "Saving file", MessageBoxButtons.OK, MessageBoxIcon.Error); } } // buttonSaveOriginalData_Click

About
private void buttonAbout_Click(object sender, EventArgs e) { (new About()).ShowDialog();

IMAGE COMPRESSION

77

} // buttonAbout_Click

Method 2 Harr Discrete Method Title Form

Public Class Form1 Dim i As Integer Private Sub Timer1_Tick(ByVal sender As System.Object, ByVal e As System.EventArgs) Handles Timer1.Tick i=i+1 If i > 25 Then Timer1.Enabled = False Dim x As New ImageCompression x.Show() Me.Hide() End If

End Sub

Open Image
Public Class ImageCompression Private Sub OpenImageToolStripMenuItem_Click(ByVal sender As System.Object, ByVal e As System.EventArgs) Handles OpenImageToolStripMenuItem.Click If ofdOriginal.ShowDialog() = Windows.Forms.DialogResult.OK Then Try picOriginal.Image = LoadImage(ofdOriginal.FileName) picNew.Image = Nothing txtCompressionLevel.Clear() Dim file_info As New FileInfo(ofdOriginal.FileName) txtOriginalSize.Text = FormatBytes(file_info.Length) txtDesiredSize.Text = FormatBytes(file_info.Length * 0.5) txtCompressedSize.Clear()

IMAGE COMPRESSION
txtCompressionLevel.Clear() Catch ex As Exception MessageBox.Show("Error loading picture file " & _ ofdOriginal.FileName & vbCrLf & ex.Message, _ "File Load Error", _ MessageBoxButtons.OK, _ MessageBoxIcon.Error) End Try End If End Sub

78

Load Image
Private Function LoadImage(ByVal file_name As String) As Bitmap Using bm As New Bitmap(file_name) Dim bm2 As New Bitmap(bm) Return bm2 End Using End Function

Compression
Private Sub HeaderDiscreateMethodToolStripMenuItem_Click(ByVal sender As System.Object, ByVal e As System.EventArgs) Handles HeaderDiscreateMethodToolStripMenuItem.Click Try Dim desired_size As Double = FormattedBytesToBytes(txtDesiredSize.Text) If desired_size < 10 Then MessageBox.Show("Invalid desired size.", _ "Invalid Size", MessageBoxButtons.OK, _ MessageBoxIcon.Error) txtDesiredSize.Focus() Exit Sub End If If sfdJpeg.ShowDialog() = Windows.Forms.DialogResult.OK Then picNew.Image = Nothing

IMAGE COMPRESSION
txtCompressionLevel.Clear() txtCompressedSize.Clear() Dim file_name As String = sfdJpeg.FileName Dim file_size As Long For compression_level As Integer = 100 To 10 Step -1 Dim memory_stream As MemoryStream = SaveJpegIntoStream(picOriginal.Image, compression_level) file_size = memory_stream.Length If file_size <= desired_size The My.Computer.FileSystem.WriteAllBytes(file_name, memory_stream.ToArray(), False) picNew.Image = LoadImage(file_name) txtCompressionLevel.Text = compression_level txtCompressedSize.Text = FormatBytes(file_size) Exit For End If Next compression_level End If Catch ex As Exception MessageBox.Show("Error saving file." & _ vbCrLf & ex.Message, "Save Error", _ MessageBoxButtons.OK, MessageBoxIcon.Error) End Try End Sub

79

JPEG Conversion
Imports System.Drawing.Imaging Imports System.IO Module JpegCompression Private Function GetEncoderInfo(ByVal mimeType As String) As ImageCodecInfo Dim HeaderDescrete As ImageCodecInfo() HeaderDescrete = ImageCodecInfo.GetImageEncoders() For i As Integer = 0 To HeaderDescrete.Length If HeaderDescrete(i).MimeType = mimeType Then Return HeaderDescrete(i) End If Next i Return Nothing

IMAGE COMPRESSION
End Function

80

Byte Conversion
Module ByteSizes Private Const ONE_KB As Double = 1024 Private Const ONE_MB As Double = ONE_KB * 1024 Private Const ONE_GB As Double = ONE_MB * 1024 Private Const ONE_TB As Double = ONE_GB * 1024 Private Const ONE_PB As Double = ONE_TB * 1024 Private Const ONE_EB As Double = ONE_PB * 1024 Private Const ONE_ZB As Double = ONE_EB * 1024 Private Const ONE_YB As Double = ONE_ZB * 1024 Public Function FormatBytes(ByVal num_bytes As Double) As String If num_bytes <= 1023 Then Return Format$(num_bytes, "0") & " bytes" ElseIf num_bytes <= ONE_KB * 1023 Then Return ThreeNonZeroDigits(num_bytes / ONE_KB) & " " & "KB" ElseIf num_bytes <= ONE_MB * 1023 Then Return ThreeNonZeroDigits(num_bytes / ONE_MB) & " " & "MB" ElseIf num_bytes <= ONE_GB * 1023 Then Return ThreeNonZeroDigits(num_bytes / ONE_GB) & " " & "GB" ElseIf num_bytes <= ONE_TB * 1023 Then Return ThreeNonZeroDigits(num_bytes / ONE_TB) & " " & "TB" ElseIf num_bytes <= ONE_PB * 1023 The Return ThreeNonZeroDigits(num_bytes / ONE_PB) & " " & "PB" ElseIf num_bytes <= ONE_EB * 1023 Then Return ThreeNonZeroDigits(num_bytes / ONE_EB) & " " & "EB" ElseIf num_bytes <= ONE_ZB * 1023 Then Return ThreeNonZeroDigits(num_bytes / ONE_ZB) & " " & "ZB" Else Return ThreeNonZeroDigits(num_bytes / ONE_YB) & " " & "YB" End If End Function Private Function ThreeNonZeroDigits(ByVal value As Double) As String If value >= 100 Then Return Format$(CInt(value))

IMAGE COMPRESSION
ElseIf value >= 10 Then Return Format$(value, "0.0") Else Return Format$(value, "0.00") End If End Function Public Function FormattedBytesToBytes(ByVal txt As String) As Double Dim base_value As Double = Val(txt) Dim ending As String = txt.Trim(New Char() {"0"c, "1"c, "2"c, "3"c, "4"c, "5"c, "6"c, "7"c, "8"c, "9"c, "."c, " "c}) ending = ending.Trim().ToLower() Select Case ending Case "bytes", "byte", "b" Return base_value Case "kb" Return base_value * ONE_KB Case "mb" Return base_value * ONE_MB Case "gb" Return base_value * ONE_GB Case "tb" Return base_value * ONE_TB Case "pb" Return base_value * ONE_PB Case "eb" Return base_value * ONE_EB Case "zb" Return base_value * ONE_ZB Case "yb" Return base_value * ONE_YB Case Else Throw New ArgumentException("Invalid size " & txt) End Select End Function End Module

81

Saving

IMAGE COMPRESSION
Public Sub SaveCompressedJpeg(ByVal image As Image, ByVal file_name As String, ByVal compression_level As Long) If compression_level < 10 Then Throw New ArgumentException("Compression level must be between 10 and 100") End If Dim encoder_params As EncoderParameters = New EncoderParameters(1) encoder_params.Param(0) = New EncoderParameter(Encoder.Quality, compression_level) Dim image_codec_info As ImageCodecInfo = GetEncoderInfo("image/jpeg") Dim mem As New MemoryStream() image.Save(mem, image_codec_info, encoder_params) image.Save(file_name, image_codec_info, encoder_params) End Sub

82

Public Function SaveJpegIntoStream(ByVal image As Image, ByVal compression_level As Long) As MemoryStream If compression_level < 10 Then Throw New ArgumentException("Compression level must be between 10 and 100") End If Dim encoder_params As EncoderParameters = New EncoderParameters(1) encoder_params.Param(0) = New EncoderParameter(Encoder.Quality, compression_level) Dim image_codec_info As ImageCodecInfo = GetEncoderInfo("image/jpeg") Dim memory_stream As New MemoryStream() image.Save(memory_stream, image_codec_info, encoder_params) Return memory_stream End Function End Module

Exit
Private Sub ExitToolStripMenuItem2_Click(ByVal sender As System.Object, ByVal e As System.EventArgs) Handles ExitToolStripMenuItem2.Click Application.Exit() End Sub

IMAGE COMPRESSION

83

APPENDIX C
Abbreviations PSNR MSE MOS PQS DPCM ASCII CLR VQ MSIL VB JIT CLS COBOL CR GIF JPEG PNG JBIG DFD : : : : : : : : : : : : : : : : : : Peak Signal-to-Noise Ratio Mean Squared Error Mean Opinion Score Picture Quality Scale : Differential Pulse Code Modulation

American Standard Code Information Interchange Common Language Runtime Vector Quantization Microsoft Intermediate Language Visual Basic Just In Time Common Language Specification Common Business Oriented Language Compression Ratio Graphics Interchange Format Joint Photographic Experts Group Portable Network Graphics Joint Bi-level Image Experts Group Data Flow Diagram

IMAGE COMPRESSION

84

APPENDIX D

References Books 1. ASP.NET The Complete Reference by Mathew MacDonald published by Tata McGraw-Hill 2. Digital Image Processing (Second Edition) by Richard.E.Woods, Rafael.C.Ganzalez published by Person Education 3. Image Processing (Theory,Algorithms and Architectures) by Maher.A.SaidAhmed published by McGraw-Hill 4. System Analysis & Design(4 edition) by Lee published by O'Reily Media 5. Programming ASP.NET by Jesse Liberty published by O'Reily Media

Websites 1. www.codingproject.com 2. www.aspfree.com 3. www.w3schools.com 4. www1.fatbrain.com 5. www.huffman.com 6. www.harrdiscrete.com