基于流行排序的多示例图像检索方法  

Manifold ranking approach to multi-instance image retrieval

作　　者：虎晓红[1,2] 钱旭[2] 王珂[3] 

机构地区：[1]河南农业大学信管学院,河南郑州450002 [2]中国矿业大学机电与信息工程学院,北京100083 [3]首都医科大学生物医学工程学院,北京100069

出　　处：《计算机工程与设计》2009年第21期5007-5009,共3页Computer Engineering and Design

基　　金：教育部科研重点基金项目(107021)

摘　　要：为了提高图像检索的性能,提出了一种基于流行排序的多示例图像检索方法,将分割后的图像表示为多示例的形式,通过给出适合图像在包空间的度量方式,有效结合流行排序和多示例学习的方法来进行图像检索。实验结果表明,采用所提出的方法的检索结果与传统的检索方法相比,检索率得到了明显的提高,检索结果更符合人的视觉习惯。To improve the performance of image retrieval result, a multi-instance retrieval approach based on manifold ranking is proposed. The segmented image is represented under multi-instance learning framework. A novel graph based learning approach to image retrieval Using multiple instances is presented, which extends the conventional manifold ranking to multi-instance Setting by introducing the adaptive geometric relationship between two bags of instances. The experiments over Corel images have shown that this approach outperforms other methods and is effective for image retrieval,

关 键 词：图像检索 流行排序 多示例学习 HAUSDORFF距离 EMD距离 

分 类 号：TN911.73[电子电信—通信与信息系统]