实现图像感兴趣区域渐进编码的新方法——一种改进的嵌入式零树小波编码算法

在分析了Shapiro的嵌入式零树小波编码算法的基础上，针对其效率的不足提出了一种改进方法。本算法通过识别重要子带，而不是EZW中的重要系数，大大减少了需要编码的零树的数量，因而节省了时间。在此改进的基础上，结合坐标数据压缩算法实现了对图像感兴趣区域的渐进编码。该算法可在接收到图像的全部编码数据之前，首先实现感光趣区域的高品质重建，且编码效率也有明显提高。

A new method for the realization of progressive coding of interested image area-an improved embedded zero tree wavelet coding algorithm

This paper described the Shapiro's embed ded zerotree wavelet(EZW)coding algorithm first.Aim-ing at the cod ing efficiency defects in EZW,an improved algorithm was proposed.By searching the significant sub-band instead of the signifi cant coefficients as in EZW,this method decreased the number of ze rotrees that should be coded.And the time con sume was saved.Based on the improved algo rithm,a simple coordi nate data compres-sion(CDC)was used to efficiently en code the pro gressive informa tion of region of interest(ROI).The pro gressive ROI cod ing technique can reconstructs the ROI im age with higher quality without re ceiving the entire image and im proves the efficiency of coding.