一种基于人眼视觉特性的零树小波编码方法

嵌入式零树小波编码被认为是目前国际上最先进的图像编码方法之一，仍然有不足之处，其中之一是该算法在编“孤立零”码时会造成比特位冗余，因此在分析了零树编码不足的基础上，提出了基于人眼视觉特性的零树小波编码方法。该方法首先对小波系数进行基于人眼视觉特性的标量量化，然后再进行零树编码。实验结果表明该方法在保证恢复图像具有良好的视觉效果的前提下，进一步提高了压缩比。

Zerotree wavelet coding algorithm based on human visual system

Embedded zerotree wavelet(EZW) image coding algorithm is one of the most advan ced image coding technologies in the world. But it has some deficiencies, one of which is that the vision character isn 't considered in coding. In this paper, a zerotree wavelet coding algorithm base d on human visual system was proposed, in which a scalar quantization based on v arying threshold was first adopted on the wavelet transform coefficients, and th e successive-approximation quantization and EZW algorithm were used in order to p roduce an embedded bit steam. The experiment results show that this image coding performs better than that of the EZW in aspects of compression ratio and image quality.