Image Compression Using Wavelet Transform Based SPIHT Algorithm a nd Performance Analysis by Changing Parameter Simulated with QccPack

Abstract

The study and analysis of wavelet image compression techniques focuses mainly ondeveloping energy ecient, computation ecient and adaptive image compressiontechnique, as the images will be a large part of future wireless data. Digital imagesare widely used in computer applications and multimedia applications. Uncompressedimage required considerable storage capacity and transmission bandwidth. Set Parti-tioning In Hierarchical Tree (SPIHT) which uses wavelet transforms, represents a veryeective form of entropy coding. The combination of integer lifting wavelet transformwith SPIHT algorithm has been widely used in the eld of image compression.The thesis conveys a detail study and simulation of advance image compression algo-rithm named SPIHT, and modication of algorithm for more compression and goodimage quality after reconstruction.In rst part of work describes in brief about image compression and various com-pression techniques. Second part gives information about wavelet Filters, how DWTis implemented for 2D. The detail about SPIHT coding/decoding algorithm is de-scribed. This algorithm simulated with QccPack library and performance analysisis done by varying various parameters of SPIHT algorithm. Two modications inSPIHT algorithm is done. First one is to select the threshold which is data depen-dent and second one is to make partitions adaptive that reduce amount of bit requirethan ordinary SPIHT.Results provide a good reference for application developers to choose the wavelet,number of decomposition level and compression ratio with option to use arithmeticcode or not for their application and modify algorithm as per their requirement.