摘要
提出了一种鲁棒的图像检索方法,它能够在检索大规模图像库时实现基于内容的快速匹配。匹配的索引来源于重量化HSV颜色空间得到的灰度图(Z值图)的区域特征。由于Z值图由反映色彩的聚类区域组成,并且使用矩方法来表达特征,因此该检索方法是鲁棒的,具有2维平移、尺度、旋转不变性,同时可以避免逐点匹配产生的时间耗费。
In this paper we propose a robust color image retrieval approach which can realize fast matching in CBIR （Context-Based Image Retrieval） when searching in large image databases.Indexes root in region features of Zimage that is the result of re-quantization in HSV color space.As Z image is made up of many color clustering regions and method of moments is used for feature representation, our approach is robust, and of 2D invariance in translation, scale and rotation, while time consumption pixel by pixel can be avoided.
出处
《测绘学院学报》
北大核心
2005年第3期194-196,共3页
Journal of Institute of Surveying and Mapping
关键词
量化
聚类
鲁棒性
图像检索
quantization
clustering
robustness
image retrieval
