基于直方图相交匹配函数的图像压缩评测准则  

A Criterion for Measuring Performance of Compressed Images Based on Histogram Cross Matching Function

作　　者：瞿晓[1] 

机构地区：[1]浙江科技学院信息与电气工程学系,杭州310012

出　　处：《科技通报》2006年第1期119-121,共3页Bulletin of Science and Technology

摘　　要：常用的基于均方根误差(RMSE)图像保真度准则不能准确地放映一些灰度图像主要敏感细节。本文提出了利用灰度直方图的相似度作为评价重建图像质量的准则。利用JPEG方法对图像进行压缩,并采用直方图相交匹配函数求得图像灰度级的匹配程度。实验结果表明,直方图相交匹配函数是一种有效的图像压缩评测准则。The popular fidelity measurement method based on root mean squared error（RMSE） is unable to completely reflect the details of the sensitive information of compressed grayscale images. A new fidelity measurement criterion has been presented, which is based on the similarity degree of grayseale histogram between the original image and compressed image. The JPEG compress method was applied, and the histogram cross matching function was used to measure the matching degree in this research. The experimental results show that the histogram cross matching function is a useful criterion for measuring the performance of compressed images.

关 键 词：友度图像 图像压缩 均方误差 直方图 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]