基于图像信息熵的四叉树检索算法

Quadtree Image Retrieval Based on Image Entropy

在基于内容的图像检索系统中,图像的抽象描述和特征提取是要解决的关键问题之一.本文提出了一种基于图像信息熵的四叉树索引结构及检索算法,可以利用较低维数的特征实现高速有效的图像检索.该方法适用于查准率要求不高情况下的相关图像检索或大型图像库的预检索.

In content- based image retrieval system, describing and extracting image＇s feature is a key step. This paper presents a new retrieval method using image＇s features described by quadtree, which is an effective method using low dimension feature. It is especially fit for retrieval in the case of lower recall rate or pre - retrieval in large scale image library.