分块自适应零树压缩编码方法研究  

Coding Method of Block Self-adapting Zero-tree Compression

作　　者：李庄[1] 陶洪久[1] 

机构地区：[1]武汉理工大学,武汉430070

出　　处：《交通与计算机》2006年第2期128-130,共3页Computer and Communications

摘　　要：基于模型和特征的编码方法是现代编码方法中的一个重要方向。在研究了图像的重要特征——图像边缘的基础上,提出了一种改进的零树编码方法——基于整数小波变换的分块自适应零树编码方法。这种编码方法主要包括:对原始图像进行分块、滤波;对最低频子带系数单独编码;对各块图像自适应地进行阈值设置;设立标志矩阵;解码后进行滤波消除分块效应。不仅加快了编码速度,而且提高了图像的峰值信噪比(PSNR)。该方法也能产生嵌入式码流,并支持多码率解码。Coding method based on model and character is an important field nowadays. The authors present an improved zero-tree coding method through researching on the image margin, which is an adaptive zero-tree coding method based on integral wavelet transform. The method includes split originality image region, filter, separately coding the lowest frequency sub-band coefficient, adaptively setting threshold of different image region, setting up flag matrix, and filtering after decoding to eliminate domino effect of split. This method not only quickens the coding speed, but also improves the peak value Signal-to-Noise Ratio（PSNR）. This method can also produce embedded data stream, and support large amount of data decoding.

关 键 词：小波变换 自适应零树 图象压缩 数据压缩 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论]