Abstract: The image retrieval system based on the global characteristics couldn摧t meet the needs of users ． In order to
overcome this shortcoming ，the image retrieval algorithm based on the user regions of interest was proposed ．This algorithm
based on the principle of discriminant saliency ，could extract the user regions of interest effectively without segmentation ．
Experimental results show this method is easily manipulated ，and has better performance than image retrieval based on global
characteristics ．

Key words: Region of interest , Salient point , Image retrieval

摘要： 针对全局特征的图像检索不能很好地满足用户的意图,提出了一种新的基于用户感兴趣区域的图像检索算法。此算法基于显著判别原则,无需对图像进行复杂的分割就能有效地提取用户感兴趣区域的显著点。实验结果表明,此算法不仅检索过程简单,在检索精度上也明显优于基于全局特征的图像检索方法。

关键词: 感兴趣区域 , 显著点 , 图像检索