摘要： 提出一种利用兴趣点进行局部特征提取及检索的新方法。该方法利用图像中兴趣点的位置信息，提取兴趣点周围局部区域的颜色及纹理特征作为刻画图像内容的主要特征，结合兴趣点的空间分布对相似图像进行检索。对真实图像数据库的检索实验表明，该方法与其他基于兴趣点的方法相比，平均检索准确率提高了11%，能够更准确地查找到用户所需的图像。

关键词: 兴趣点, 局部特征, 局部二值模式

Abstract: This paper presents a novel local feature detection and image retrieval method based on interest points. The method utilizes the location information of the interest points, and extracts the local color feature and local texture feature to express the contents of an image. Experimental results show that the method improves the average retrieval accuracy by 11 percent, compared with other interest points based retrieval method.

Key words: interest points, local features, local binary pattern

中图分类号: 

• TP317.4