基于分量分析的图像内容检索的实现

本文基于图像内容检索中,常用的全图的颜色直方图算法过于简单的缺点,提出了一个改进,即将图像进行了分割.形成若干子块,这样就提供了一定程度的位置信息,而且可以对含用户感兴趣的子块加大权重,提高检索的查询智能性和查准查全率.

Realization of Content Based Image Retrieval Using Principal Component Analysis

This article based on the image content retrieval,the commonly used total graph's color histogram algorithm too simple shortcoming,proposed an improvement,that the image has carried on the division,forms certain subblocks,has like this supplied the certain extent positional information,moreover may to contain the user to be interested the subblock to enlarge the weight,enhances the retrieval the inquiry intelligence and looks up the accurate recall