一种改进的嵌入零树小波编码算法RA-EZW

Study of RA-EZW-based image coding algorithm

针对嵌入零树小波编码算法(EZW)中不同阈值下相同系数被多次扫描编码的不足,通过对小波变换后系数幅值的变换调整,减少孤立零点,增加零树的数量,从而提高编码效率.实验结果表明:基于幅值变换的嵌入零树小波编码算法在信噪比、编/解码时间、图像恢复质量等方面都高于EZW算法,在恢复质量方面,比特率较低时效果尤为明显.

RA-EZW（embedded zerotree wavelet based range-adjusting） was presented on the basis of EZW＇s deficiency. The method decreased the coding numbers of the same coefficient under the different threshold by range-adjusting to improve the coding efficiency. In comparison with the previous EZW algorithm, the experiment results showed that the new scheme performed better in compression ratio, image quality, coding speed and coding efficiency, especially in the condition of low bit-ratio.