Lossless and near lossless Compression approaches for gray Level images

Abstract

The standard objective for any image compression algorithm is to newlineincrease the compression ratio However with the advent of current medical newlineimaging modalities the problem of image compression deviates from the newlineoriginal objective The compression is no more a method to increase he newlinecompression ratio alone but additional factors such as the complexity of the newlinealgorithm and the scalability are also considered In this thesis new image newlinecompression methodologies are proposed for the lossless and the near lossless newlinecompression of gray level images for improvising the compression ratio At newlinethe same time the complexity of the algorithm and the scalability feature are newlinegiven consideration newlineTwo new approaches are proposed in this thesis for near lossless newlinecompression of the images In the first approach the input image is initially newlinevisually quantized to generate a pre quantised image The visual quantisation newlineprocess generates the near lossless condition The pre quantised image is newlineencoded using a block based lossless DPCM followed by an entropy newlineencoding In the second approach the input image is pre processed using a newlinelogarithmic transformation method which introduces the near lossless newlinecondition After the transformation the resultant image is encoded using the newlineinteger wavelet transform and entropy encoded For both the methods the newlineencoders are tested with natural and medical images and the results are newlinecompared with the CALIC or the JPEG LS newline newline