一种基于视觉特性的小波包图像编码压缩方法

A Method of Wavelet Packet-Based Image Coding Based on Human Visual System

针对图像能量分布特点及人眼对图像特殊区域敏感程度的不同,提出利用小波包对图像的低频子带和高频子带都进行多分辨率分解,并充分利用人眼的视觉特性和不同子带间小波包系数的相关性对系数进行编码。实验结果表明在保证优先编码视觉上最重要系数的同时,能够进一步提高图像复原质量。

In view of the image energy distribution characteristic and the difference of human eye to the image special region sensitive degree, using the wavelet packet carries on the multi-resolution to the image low frequency subband and the high frequency subband to decompose is proposed, and by making the best of human eye＇s visual characteristic and between the different innertube the wavelet packet coefficient＇s relevance. The experiment results indicated that while assuring the most crucial coefficient on the vision of priority encoding, it could improve the image restoration quality as well.