【摘要】 为了弥补全局特征在描述图像内容上的不足,论文描述了一种基于目标区域的图像检索方法:从用户指定的区域中分割并提取出所包含的目标,然后利用目标区域的视觉特征进行图像检索。在图像检索时,提取目标区域对应的子图像特征代替全局图像特征进行图像相似性匹配。实验结果表明,基于目标区域的图像检索效果比基于全局图像特征的检索效果有较好的改善。更多还原

【Abstract】 To overcome the disadvantage of describing contents of an image only with global features,a color image retrieval algorithm based on object regions is proposed.Firstly user selects an approximate area with a rectangle,then object regions are extracted from it.During the image retrieval,the vision features of sub images,which the object regions correspond to,are extracted.Finally,these features are used in image matching in place of global image features. Experiment results show that the color image retrieval based on object regions is superior to that based on global image features and more suitable for human vision.更多还原