综合距离和相关性的图像检索算法(英文)

Combined distance and correlation measurement for content-based image retrieval

在基于内容的图像检索中,大多数都是采用距离来测试两幅图像的相似性。提出了一种新的计算相关系数的方法并结合这种方法和距离来判断两幅图像的相似性,将其应用于CBIR(content-based image retrieval)系统。在对所提出的算法进行的实验中,用了10000幅图像来测试了所提出的算法,实验结果表明:在同一个CBIR系统中,引入相关性能够提高图像的检索精度,解决了只用距离来判断两幅图像相似性的不足,对于基于低级特征的图像检索系统是一个很好的改进。

While existing CBIR （content-based image retrieval） system primarily use distance to judge whether two images are similar or not, an additional correlation test to formulate a combined content similarity measurement is presented. To test and evaluate the proposed algorithms, the experiments with a database of lO 000 images are made, the experimental results show that, with the same content-based image retrieval algorithm, adding correlation to the existing operational process can improve the performance of contentbased image retrieval significantly, providing a possible solution to the problem that minimum distance in low-level features does not necessarily lead to a content match.