关键词: 图像检索, 二进小波变换, 边缘检测, 颜色特征, 形状特征

Abstract: Because of the wavelet transform has good local characteristics and multi-scale characteristic, it can be  multi-scale edge approximation, this makes it wide range of applications in image singularity detection and feature extraction. Using dyadic wavelet transform on edge detection, use of edge gradient direction histogram express shape features, use of HSV color model express image color features, this paper puts forward an algorithm that image retrieval approach combining color feature and shape feature. The experimental results show that the algorithm not only has a strong robustness to the image illumination change, and geometry change, and has a good retrieval performance.

Key words: image retrieval, dyadic wavelet transform, edge detection, color feature, shape feature