A Study Of Still Image Compression And Coding Based On Zero-tree Wavelet

With the advent of the era of information and the development of computer networktechnology, people need to storage, processing and transmission of information isbecoming more and more images as media and means of passing information is veryimportant. To explore efficie nt image compression coding method has becomeinternationally recognized as an important research hotspot. Wavelet theory is a newbranch of mathematics, developed in recent years because of its good spatial frequency ofa local characteristic and consistent with human visual characteristic of transformationmechanism, become one of the main research direction of image compression coding.This paper first introduces the basic theory of image compression coding andevaluation index of digital image compression and then elaborated based on wavelettransform image compression coding technology, probes into the wavelet coding steps,several key issues on wavelet coding, the advantages and disadvantages of wavelettransform coding. By the way of analyzing the experimental data, the distributioncharacteristics of images after wavelet transform coefficient of energy, and then illustratesthe advantages of wavelet transform for image compression coding.Secondly, this paper deeply analyzes based on embedded zerotree wavelet codingmethod, expounds the main characteristics of EZW algorithm, principle and realizationprocess of the algorithm, including zero tree structure, the wavelet coefficient of thescanning mode, embedded zerotree wavelet coding method of algorithm process, thealgorithm of improved algorithm is proposed combined with the human visualcharacteristics.Improved algorithm mainly aimes at the important factor for the edgeinformation,so that defined as the reclassified the important coefficient;According to findthe sons of maximum to judge zero tree roots or isolated zero;The coding phase of finallyabandoned the arithmetic coding way, the harvard man encoding encoding, to improve thecoding efficiency. Finally through the Matlab simulation experiment, proves that the improvedembedded zerotree wavelet coding algorithm has higher peak signal to noise ratio andbetter visual effect, so that the improved algorithm is feasible and effective.