Image Compression Using Wavelet Transform

1 Abstract The wavelet transform shows a lot of promise in image compression. Very high compression rates are achieved by using wavelet transform prior to quan-tization and entropy coding. This report discusses how wavelets can be used in image compression. This report handles the problem from an engineer's perspective. I am conndent that this report can be used for writing a primitive wavelet based image coder. 2 Introduction Digital images contain huge amount of data, for example a 512x512, 24 bit true-color image require 512x512x3 bytes. So to decrease the memory needed to store the image and to reduce the bandwidth requirement to transmit the image, it is essential to compress it as eeciently as possible. Broadly compression techniques can be divided in to two categories. lossy compression loss-less compression In loss-less compression, the compression rate is low but the actual image can be reconstructed perfectly from the compressed version (Well there will be some error because computer can not handle real numbers), one example of such an algorithm is run-length encoding. On the other hand, in lossy compression , we sacriice the perfect reconstruction of the image in order to have better compression rates. Ofcourse, in doing so we introduce some distortion in the image. The aim in lossy compression is to have maximum compression given a distortion. The examples of such algorithms are transform coding and subband coding etc. Redundancies in the image are exploited to compress it. Two types of redundancies are encountered in images.