优先编码重要系数的邻域的零树小波图像压缩算法

Zerotree wavelets coding algorithm by preferentially encoding the neighbor of significant coefficients

在EZW算法的基础上提出了一种优先编码重要系数的邻域的零树小波图像压缩算法.因为在较大阈值下重要的小波系数的邻域的系数在当前阈值下重要的概率很高,于是优先扫描这些系数,并且全部进行量化编码,即使这些系数不重要.实验表明,在相同压缩比下,该方法能显著提高主观视觉感受和PSNR.

This paper presents an improved embedded zerotree wavelets coding algorithm by preferentially encoding the neighbor of significant coefficients. Because the coefficients around the significant wavelet coefficients in the larger threshold are very likely significant in the current threshold, we preferentially scan these coefficients and all of them are refined even they are not significant. The experimental results show that the proposed method can significantly improve the subjective visual experience and PSNR.