摘要
本文简单介绍了小波变换编码的流程、特点和小波基的选取 .并根据灰度图像经小波变换以后系数的分布、统计特性 ,对现今十分流行的嵌入零树 (EZW)、分层树集合分割排序 (SPIHT)算术编码方法进行研究 .综合分析了该类算法的优缺点 ,并提出 2种不影响算法整体结构的改进意见 ,对SPIHT编码算法增加了一些初始化操作 .实验结果表明 ,在一定比特率下 。
This paper gives a brief introduction about the process of image coding, features and the choosing of wavelet filters.Based on the good statistical feature of wavelet image, it presents a study of the most popular coding methods: embedded zero wavelet (EZW), set partitioning in hierarchical trees (SPIHT). Synthesizing the advantages and disadvantages of this kind of methods, SPIHT is developed by two ways which don't affect its main frame to add some initial treatments. Results show that the improved algorithm has better effects in certain applications than the unimproved one.
出处
《东南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2003年第z1期82-85,共4页
Journal of Southeast University：Natural Science Edition
关键词
零树
嵌入零树
分层树集合排序
不重要系数表
重要系数表
不重要集合表
zero tree
embedded zero wavelet
set partitioning in hierarchical trees
the list of insignificant pixels
the list of significant pixels
the list of insignificant sets
