论文部分内容阅读
小波变换具有良好的空间 -频率局部化性能 ,主要表现在频率压缩特性、空间压缩特性、系数分布的相似性 3个方面 ,这些特性都有利于进行图象压缩 .但是早期的小波压缩算法大多没有利用系数分布的相似性 .该文借鉴了零树算法和 Rinaldo块预测的思想 ,提出了一种新的旨在压缩重要小波系数结构性冗余的静止图象压缩方法 ,实验结果证明了这种方法的有效性 .
Wavelet transform has good performance of space-frequency localization, mainly in the three aspects of frequency compression, space compression and coefficient distribution, which are good for image compression.But most of the early wavelet compression algorithms do not have Using the similarity of coefficient distribution.This paper draws on the idea of zerotree algorithm and Rinaldo block prediction and proposes a new method for compressing still images which is used to compress the structural redundancy of important wavelet coefficients.Experimental results show that this The effectiveness of the method.