Interesting Region Compression Coding Technology Research Based On Wavelet Transformation

1. IntroductionAs information technology, computing technology, modern communications technology, microelectronics and network technology has developed rapidly, people has entered the information age today, the image information processing, storage and transmission will become more and more important in the social life, it was right to accept the demands of image information is increasingly urgent. The image usually originates from the natural scenery, its primitive shape is the simulation quantity which changes continuously. However, image is processed and transmited by digital form, because its quality is good, the cost is low, the reliability is high; Could relay transmission and could be copied many times, and cannot create the noise and the nonlinear distortion accumulation; It is easier to encrypt, realize by the VLSI chip, and connect with the computer, etc. It has made digital image processing already to become current and the future main trend of development in this field. But because the data of image information is too huge, it is restricting the storage and transmission of image information. So the problem that the memory is too huge when storage and the channel band is too broad when transmission is appeared. Therefore, the image coding and the compression is becoming more and more important both in application and theoretics as key technologies, and always is one research branches of the information processing field which is most active.This thesis analysis the prensent developmental situation of the image compression coding technology, choses object-oriented code technology as research content based on the wavelet transformation, proposes one kind of improved Bitplanes shift algorithm.2. Analysis of wavelet theoryThe Wavelet Transform coding is a kind of coding method which is proposed along with the research of Wavelet Transform theory in recent years. It must have bright future. The Wavelet Transform has very good time-frequency or spatial- frequency character and multi- differentiation, So it is very suited the image compression coding project which is designed according to human vision system character, also is extremely advantageous in the image layered transmission. So Wavelet Transform theory is becoming more and more important in image compression coding theory. JPEG2000 and MPEG4 which are the up to date image compression standard both adopt Wavelet Transform.The general work of wavelet image coding is to select a fixed wavelet base, make the wavelet decompose to the image, study the reasonable quantification plan, the scanning way and the entropy encoding method in wavelet region. The key problem is that we should propose a effective scheme according to the character of wavelet transformation region. In the stationary picture wavelet coding domain, one of the most effective methods is the wavelet zero tree coding method, and other methods which are based on this method, they all fully use the range relativity of different scale layer's coefficient to forecast in the wavelet region, it has solved the zero position coding problem well. The zero wavelet coefficient has zero tree structure, so we could use a zero tree root symbol to figure the zero coefficient of abundant tree structure. Thus, the coding efficiency is more than JPEG which is based on DCT transformation. In recent years, many new improved arithmetics which is based on EZW have emerged, such as SPIHT, SPEDK, etc. EZW and SPIHT is one of the best wavelet encoder so far. This thesis adopts EZW.3. Region of interest codingImage compression based on Lossless or nearly Lossless Region of Interest (ROI ),means to com-press interesting regions in an image without loss,and to compress uninteresting regions with loss.Thus,desired high fidelity image information is acquired while remaining a high compress rate.This technology solves the dif-ference between the image quality and the compress rate. In this paper, I use two ROI coding method: MAXSHIFT method and Scaling-based method. They are all realized by placing a lower cofficient on the rest of the image (BG) than the region of ROI.4. Improvement of two existing ROI coding methodThe two kinds of coefficient shift ROI algorithm are all deficient.　 MAXSHIFT method definition is not selected random and it is not fit for multi–ROI-image.Scaling-based method is coding for the figure of ROI image and the coding efficiency is low. For the reason above, I bring forward a improved algorithm. In this improved method, the most significant bitplanes of ROI are shifted but not all of the bitplanes like the standard described. By emulating, the improvement of ROI coding algorithm is effectiver.5. ConclusionIn this treatise,with the basic study of current domestic and foreign's two normal ROI coding algorithm(MAXSHIFT method and Scaling-based method) ,I introduce a improved ROI coding algorithm and realize it by MATLAB prgraming.There are four basic predominances in improved ROI coding algorithm: (1) It is fit for any figure of ROI image and it don not need the information of the figure of the image. (2)It allow different wavelet subband have different ROI definition. (3)It use suitable shift coefficient to control the image quality contrast of ROI region and GB. (4) In the condition of low bitrate, it can code many ROI region with different priorities efficiently in a image. The combination of the result of the paper and JPEG2000 will have a abroad appliance foreground.