Image Compression Algorithms using Wavelets: a review Sunny Arora Department of Computer Science Engineering Guru PremSukh Memorial college of engineering City, Delhi, India Kavita Rathi Department of Computer Science Engineering DeenbandhuChhotu Ram University of Science and Technology, Sonipat, Haryana, India Abstract – Wavelet transformation is powerful feature for the signals and frequency analysis of an image. Wavelet family emerged as an advantage over Fourier transformation or short time Fourier transformation (STFT) .Image compression not only reduces the size of image but also takes less bandwidth and time in its transmission. This paper uses two image compression algorithms SPIHT and EZW for comparing the quality of images and the quality of images is compared by taking PSNR and MSE of images. Analysis of the quality measures have been carried out to reach to a conclusion. Keywords –SPIHT, EZW, MSE, PSNR, LSP, LIP, ZTR I. INTRODUCTION Image compresssion is divided into two categories, lossless compression[1] and lossy compression. In lossless compresssion, the image is reconstructed by using image information and while lossy compression is allowed loss when the image information is reconstructed. In general, the image is required to fully reconstruct without loss. However, the lossy image is used for obtaining a higher compressed ratio. For a compressed image, the number of bits used in the compressed representation of the image divided by the number of pixels in the image is defined as bit rate achieved by the compressor, measured in bits/pixel. The image can be compressed in such a way by using quality metric or resolving several distortion inorder to match the original image in lossy compression. Peak Signal to Noise Ratio(PSNR) is a quality metric, because increasing PSNR values indicates increasing reconstructed image fidelity. Here M and N denote the image width and height, respectively; f(m,n) and g(m,n) denote original and reconstructed iamge; and B denotes the dynamic range(in bit) of the original image. SPIHT[2] [3] algorithm was introduced by Said and Pearlman[4] and is improved and extended version of Embeded Zerotree Wavelet(EZW) coding algorithm introduced by Shapiro[5][6]. Both algorithms work with tree structure, called Spatial Orientation Tree (SOT), that defines the spatial relationships among wavelet coefficients in different decomposition subbands. In this way, an efficient prediction of significance of coefficients based on significance of their parent coefficients is enabled. SPHIT is a low-complexity progressive image compressor. This enhanced implementation of a zerotree algorithm efficiently encodes zerotrees with a relatively modest level of complexity and produces an embedded bitstream. A higher image compression ratio can not be obtained by using SPHIT algorithm only, it can be obtained by using a lossy algorithm based on SPIHT encoding algorithm is proposed in this paper. Firstly, the wavelet coefficients are divided into sevral blocks, then the importance of different blocks are by adopting SPHIT algorithm of different blocks are encoded by adopting SPHIT algorithm of different bit rate, in order to improve the compression ratio. II. FRAME OF SIMULATING SYSTEM According to the importance of the wavelet coefficients blocks, the simulation system consists of three parts: a Discrete Wavelet decomposition of image data, a SPHIT encode module which performs the coding of wavelet coefficients in the block by bit rate and a bit allocate module which controls the SPHIT encode module that input the bit stream as shown in the fig.1. International Journal of Innovations in Engineering and Technology (IJIET) Vol. 4 Issue 1 June 2014 205 ISSN: 2319 – 1058