Application And Research Of Wavelet Analyse In Digital Image Compression

With the ever-growing multimedia technique, people are looking forward to new image compression techniques with better performances. Taking the storage and transfer of mass image data into consideration, not only could favorable image compression techniques achieve high compression ratio and fidelity, they should also meet the demands of web applications such as progression transmission and. Thereby, to satisfy the practical applications, image compression need further detailed study, thus having developed into a specialized research field known as image coding. Frequently used major coding methods include predicl code, transform code, statistic code, subband code, fractal code, model code, vectol quantization code, neural networks code, wavelet transform code, etc.In this paper, we concentrate on image coding method based on wavele transform. In the firt place, the basic principles and methods of image coding theory are summarized. Also, we introduce some familiar compression techniques and international compression standards on image and video frequency as well. Then, we present wavele transform ,along with the relative theories when it is applied to practical uses. At last, we focus on Embedded Zerotrees Wavelet algorithm based on wavelet transform. Through anatomizing EZW coding algorithm and taking the problems emerging during its process of modification into account, we propose an improved modification of zerotrees coding algorithm, sufficiently considering the effects of human being's vision characteristics on quality of reconstituting image. The improvement of algorithm aims at promoting compression ratio with minor loss of quality of reconstituting image, and is validated by experiments. Eventually, the improving project based on EZW coding is implemented by C++. According to the analysis of image compression result, the improved algorithm proves its rationality and validity. Another significant merit of this algorithm is that we can finish coding at any node, thus being able to adjust compression ratio according to the demand of quality of reconstituting image.