基于优化TCQ的小波域图像编码  

Optimized TCQ-Based Image Coding in Wavelet Domain

作　　者：王安红[1] 李志宏[1] 郑义[1] 王宇洋[1] 

机构地区：[1]太原科技大学电子信息工程学院,山西太原030024

出　　处：《铁道学报》2011年第8期58-62,共5页Journal of the China Railway Society

基　　金：国际科技合作计划(2010DFA11010);国家自然科学基金(61073142);山西省国际科技合作项目(2011081055);山西省青年科技研究基金(2008021020)

摘　　要：本文提出一种基于优化网格编码量化(TCQ)的小波图像编码方法。该方法首先利用小波子带间的相关性对小波系数进行改进四叉树遍历扫描,然后对扫描后的非零系数采用一种优化TCQ进行量化,最后将量化系数进行熵编码。由于优化TCQ方案利用不同子带小波系数的重要性进行不同精度的量化,因此所提出的小波图像编码方法比经典的零树编码取得更好的率失真性能,同时还具有编码复杂度适中的优点。实验验证了方案的有效性。The image coding method based on optimized trellis coded quantization（TCQ） is presented in the wavelet domain.First,the improved scanning of quadtree exploits correlation between the sub-bands to classify the zero trees and non-zero trees in wavelet coefficients,and then the non-zero tree is quantized by adopting optimized TCQ.Finally,the quantized coefficients are compressed by entropy coding.Because the optimized TCQ considers the characters of wavelet coefficients in different subbands,the proposed scheme is expected to achieve better rate-distortion performance than the classical zero-tree coding with modest encoding complexity.The experimental results confirm the effectiveness of our scheme.

关 键 词：小波变换 零树编码 四叉树 TCQ 图像编码 

分 类 号：TN919.81[电子电信—通信与信息系统]