Image Compression Coding Based On Multi-bank Wavelets Transform

The core technology of image compression coding i&the transformation of image data.The still image compression standard represented by JPEG is based on discrete cosine transform(DCT),which was highly developed both in theory and in application.The main problem of the discrete cosine transform is the im-age has block noise.The new image compression standard JPEG2000 adopts the transformation form of 2-bank wavelets transform,which greatly improves the compression performance,but high computational complexity and high cost of ac-tual hardware have cause it not widely used.To solve these problems,this paper adopts the multi-bank wavelets transfor-m to perform image compression coding,takes 16-bank wavelets transform as an example and discusses the details of compression coding.The multi-bank wavelets transform not only has the advantage of wavelets transform without block noise,but also can design it into block algorithm when the software and hardware are actually designed,which greatly reduces the hardware cost.Based on 16-bank wavelets transform,this paper proposed new quantitative strategy,scanning mod-e,and the improved QM arithmetic encoder,including the improvement of co-efficient model and the selection of probability status table,implements a set of complete and practical image compression coding scheme.Experiments show that the reconstructed image is not only better than JPEG on the objective standard-PSNR,under the subjective quality is also better than JPEG algorithm.Therefore,this method has deep research value and wide application prospects.