ITSI Transactions on Electrical and Electronics Engineering (ITSI-TEEE)
_______________________________________________________________________________________________
_______________________________________________________________________________________________
ISSN (PRINT) : 2320 – 8945, Volume -6, Issue -1-2, 2018 23
A Novel Approach for Efficient Image Compression Using PCA and DWT
1Priyanka, 2Raman Goyal
1M.Tech CSE, LLRIET, Moga, Punjab
2Assistant Professor CSE, LLRIET Moga, Punjab Email: 1[email protected], 2[email protected] Abstract-Image compression is an approach that has been
used for reducing storage capacity of the images. In the process of image compression images information has been reduced in such a way that quality of mage will not be affected and minimum storage capacity can be achieved.
The paper is based on compression with using Steganography technique. In this paper hybrid approach for image compression algorithmsnamely PCA, DWT and DCT has been used for image compression process. In this process image has been subdivided into blocks and Eigen values from each block has been computed and these coefficients has been used for encoding process that reduces size of the images. This process of compression preserve image quality during process of decompression that can be analyzed from performance evaluation parameters.
Keywords: Steganography, Image compression, DCT, JPEG, DWT.
I. INTRODUCTION
1.1 Digital Image Processing
The field of digital image processing refers to processing digital images by means of a digital computer. Digital image processing encompasses processes whose inputs and outputs are images and, in addition, encompasses processes that extract attributes from images, up to and including the recognition of individual objects. The processes of acquiring an image of the area containing the text, pre-processing that image, extracting (segmenting) the individual characters, describing the characters in a form suitable for computer processing, and recognizing those individual characters comprises of digital image processing.
1.2 Image compression
Image compression is a key technology in transmission and storage of digital images because of vast data associated with them. Image compression is important for many applications that involve huge data storage, transmission and retrieval such as for multimedia, documents, videoconferencing, and medical imaging.
Uncompressed images require considerable storage capacity and transmission bandwidth. The objective of image compression technique is to reduce redundancy of the image data in order to be able to store or transmit data in an efficient form. This results in the reduction of file size and allows more images to be stored in a given amount of disk or memory space [12]. Image compression is an application of data compaction that
can reduce the quantity of data. Image compression techniques reduce the number of bits required to represent an image by taking advantage of these redundancies. An inverse process called decoding is applied to the compressed data to get the reconstructed image. The objective of compression is to reduce the number of bits as much as possible, while keeping the resolution and the quality of the reconstructed image as close to the original image as possible.
1.2.1 Reduce the Correlation between Pixels
Why an image can be compressed? The reason is that the correlation between one pixel and its neighbor pixels is very high, or we can say that the values of onepixel and its adjacent pixels are very similar. Once the correlation between the pixels is reduced, we can take advantage of the statistical characteristics and the variable length coding theory to reduce the storage quantity. This is the most important part of the image compression algorithm; there are a lot of relevant processing methods being proposed. The best-known methods are as follows:
Predictive Coding: Predictive Coding such as DPCM (Differential Pulse Code Modulation) is a lossless coding method, which means that the decoded image and the original image have the same value for everycorresponding element.
Orthogonal Transform: Karhunen-Loeve Transform (KLT) and Discrete Cosine Transform (DCT) are the two most well-known orthogonal transforms.
The DCT-based image compression standard such as JPEG is a lossy coding method that will result in some loss of details and unrecoverable distortion.
Subband Coding: Sub-band Coding such as Discrete Wavelet Transform (DWT) is also a lossy coding method. The objective of sub-band coding is to divide the spectrum of one image into the low pass and the high pass components. JPEG 2000 is a 2-dimension DWT based image compression standard.
1.3 Need of Compression
Uncompressed images can occupy a large amount of memory in RAM and in storage media, and they can take more time to transfer from one device to another.
The need for sufficient storage space and more bandwidth because long transmission time is required
ITSI Transactions on Electrical and Electronics Engineering (ITSI-TEEE)
_______________________________________________________________________________________________
_______________________________________________________________________________________________
ISSN (PRINT) : 2320 – 8945, Volume -6, Issue -1-2, 2018 24
for uncompressed image,so there is only one solution is to compress the image.
II. METHDOLOGY
In the process of image compression images information has been reduced in such a way that quality of mage will not be affected and minimum storage capacity can be achieved. In compression method some information of original image is lost and the lost information of the image cannot be restored which affects image quality, this method is also known as lossy method.
In the purposed work image has been acidized in the system for image compression process. In the process of image compression image has been used for various operation to reduce the size of the image in such a way so that storage management can be achieved and minimum data loss to be occurred. PCA compression approach has been used for image compression to reduce size of the image so that storage capacity can be decreased to a particular extent. PCAcompression utilizes various wavelet coefficients to encode and compress image information. DWT and DCT have been used in this process to convert image to compressed dataoperations to reduce size of the image. In this process encoding has been done on the basis of different encoder schemes that have been used by arithmetic encoder to convert images wavelet coefficients.
Fig. 1 encoding model of PCA compression Fig1 shows encoding model of PCA compression. This model uses various expressions for compression of RGB image. DWT has been implemented on the image to convert image into different sub bands. Sub bands have been converted to YCBCR model for extraction of luminance3 and chrominance. After this DCT and quantization, PCA and Arithmetic coding has been done.
Fig. 2 decoding model of PCA compression Fig 2 depicts decoding model on PCA decompression algorithm. On the basis of this model image has been converted to bit streams after compression. Bit streams have been used for decompression to extract original data. For this inverse of PCA compression has been undergoes. Inverse DCT has been implemented to form original sub bands of the image. All the sub bands undergo process of inverse wavelet transformation so that image can be converted to original format from bit streams.
Discrete Wavelet Transform
Discrete Wavelet Transform (DWT) will decompose the image into four different bands. These bands are low- low (LL), low-high (LH), high-low (HL) and high-high (HH). Haar wavelet filter is used for the interpolation of the LL band. On the basis of level of decomposition LL band has been decomposed into further different levels.
This process computes the spatial and frequency domain information from the images. These information captures from the image have been under different small waves of frequency.
Discrete Cosine Transform
After computation of Low-Low band from input image DCT has been used on these components of the image to divide into blocks of 8 * 8.
Quantization
Quantization is the process that has been used for image compression to set coefficient values to a particular round figure value. DCT is a lossless procedure. The data can be precisely recovered through the IDCT. In the process of quantization each values of image matric have been divided by quantization matrix value. After division quantization coefficients have been rounded off and stored for PCAcoding.
PCA
Principal component analysis comprises various different phases for accommodationof the compression in the cooled images. In this process mean and subtracted images has been computed for extraction of data and covariance matrix and Eigen values are features of PC components of the original image that can be used for reduction process.
Original Input Image
Discrete Wavelet Transfor
m
Discrete Cosine Transform
Principal Component
Analysis Compressed
Image
Compresse d Image
PCA decompressi
on Coding
IDCT
Inverse Discrete Wavelet Transform Decompress
ed Image
ITSI Transactions on Electrical and Electronics Engineering (ITSI-TEEE)
_______________________________________________________________________________________________
_______________________________________________________________________________________________
ISSN (PRINT) : 2320 – 8945, Volume -6, Issue -1-2, 2018 25
Principal Component Analysis has been used for extraction of the features from the different images.
This approach generates 2 dimensional covariance matrixes for the image data. The co- variance matrix is used for development of Eigen values and Eigen vector.
After this process of principal component analysis coding arithmetic coding has been used for generation of image coefficients in compressed form.
III. RESULTS
In the process of image compression different standard images has been taken as input so that compression can be done. These images are Lena, Baboon and Pepper.
These images with size of 512*512 has been taken as the input images and used for compression based on proposed technique. On the basis of proposed technique image has been compressed and stored in compressed file. And compression ratio has been computed.
Fig 3 original and decompressed images This figure represents images that have been input to the system and images that have been extracted after decompression process. Quality of the images gets changes due to quantization and arithmetic coding in compression approaches. On the basis of proposed approach various parameters have been analyzed for performance evaluation of proposed system. These parameters are compression ratio, peak signal to noise ratio and structure similarity index matrix.
Fig 4: Graph for SSIM Analysis
This figure represents comparison between values of SSIM based on different compression approach at all the images that have been used in simulation process.
TABLE I: REPRESENTING COMPRESSION RATIO, SIZE AND PSNR S.N. COMPRESSION
TECHNIQUES IMAGES ORIGINAL
IMAGE SIZE COMPRESSED
IMAGE SIZE C.R. % PSNR SSIM
1 DCT+ STGO
LENA 99.60 KB 57.82 KB 42 % 32.53 0.59
PAPPER 90.51 KB 52.86 KB 42 % 35.37 0.75
BABOON 137.98 KB 82.02 KB 41 % 29.44 0.87
2 DWT+STGO
LENA 99.60 KB 48.48 KB 51 % 31.98 0.58
PAPPER 90.51 KB 49.83 KB 45 % 36.17 0.79
BABOON 137.98 KB 58.47 KB 58 % 26.69 0.82
3
DWT+PCA AND ARITHMETIC CODING
LENA 99.60 KB 37.32 KB 63 % 34.70 0.84
PAPPER 90.51 KB 37.29 KB 59 % 38.67 0.91
BABOON 137.98 KB 37.58 KB 73 % 32.95 0.99
This table represents various parameters for different approaches for compression. On the basis of these values proposed approach provide better structure
similarity index matrix and compression ratio rather than of previous approaches.
IV. CONCLUSION
In the proposed work PCA based compressions have been used for image compression. In the PCA compression various algorithms has been used for reduction of size of the image. These algorithms perform various operations to reduce size of the image that are DWT and DCT. DWT divided image into four different
sub band that are LL, LH, HL and HH. DCT divides the image into small different blocks of size 8*8. These blocks undergo the processing of quantization for reduction of high frequency coefficients of the image DCT block. PCA has been implemented on these different blocks to divide image into sub blocks and unimportant coefficients of high frequency division have been eliminated on the basis of arithmetic coding. Image a) Lena b) Pepper
c) Baboon
ITSI Transactions on Electrical and Electronics Engineering (ITSI-TEEE)
_______________________________________________________________________________________________
_______________________________________________________________________________________________
ISSN (PRINT) : 2320 – 8945, Volume -6, Issue -1-2, 2018 26
compression using SKIP run length coding that convert vector of DCT coding into groups of zeros and EOB is stated if after zero no other coefficient value has been available. On the RLE based groups Huffman coding has been implemented that uses Huffman codes from the table to convert data to bit-streams. These approaches are implemented in sequential way such that image size get reduces and effect only comes to the luminance of the image that cannot be easily predicted by naked eyes.
In the purposed work various parameter have been evaluated for performance evaluation of purposed work.
Compression ratio, Peak signal to Noise ratio and SSIM has been computed for purposed work.
REFRENCES
[1] Amir Said and William A.Pearlman,“A New Fast and Efficient Image Codec Based on Set Partitioning in Hierarchical Trees”, IEEE Transactions on Circuits and Systems for Video Technology, Vol. 6, June 1996.
[2] M. Weinberger, G. Seroussi, and G. Sapiro.
LOCOI: A low complexity, context-based, lossless image compression algorithm. Proc. Data Compression Conference, pages 140–149, March 1996.
[3] Subramanya A, “Image Compression Technique,“Potentials IEEE, Vol. 20, Issue 1, pp.
19-23, Feb-March 2001.
[4] J. Liang and T. D. Tran, “Fast multiplier less approximations of the DCT with the lifting scheme,” IEEE Trans. Signal Processing, vol. 49, pp. 3032–3044, Dec. 2001.
[5] L. Huang, M. Sciences, and N. Zealand, “Image compression based on fuzzy technique and wavelet transform,” pp. 51–56, 2002.
[6] X. Li, Y. Shen, and J. Ma, “An efficient medical image compression”, in Engineering In Medicine And Biology 27th Annual Conference, Shangai,China, IEEE, September 1-4 2005.
[7] Espen Volden, Gérard Giraudon, Marc Berthod,
“Image redundancy and classification”, Computer Analysis of Images and Patterns, Vol.
970 of the series Lecture Notes in Computer Science, pp. 206-213, June 2005.
[8] Venkata Rama Prasad , Ramesh Babu, “Adaptive Gray Level Difference to Speed up Fractal Image Compression” International Conference on Signal Processing, Communications and Networking, pp. 253-258, Feb. 2007.
[9] Xiaowen Li, Xiang XieGuolin Li, Li Zhang, Zhihua Wang, “Low-complexity near-lossless image compression method and its application- specific integrated circuit design for a wireless endoscopy capsule system” Author Affiliations J.
Electron. Imaging. 16(1), March, 2007.
[10] Rafael C. Gonzalez, Richard E. Woods, “Digital Image Processing”, published by Pearson Education, Inc, 2008.
[11] Wei-Yi Wei, “An Introduction to Image Compression”,Master’s Thesis,Graduate Institute of Communication Engineering, National Taiwan University, Taipei, Taiwan, ROC.
[12] Sonal, D. Kumar, “A Study of Various Image Compression Techniques”,Department of Computer Science & Engineering, Guru Jhambheswar University of Science and Technology, Hisar.