A new algorithm of image retrieval based on multiple-feature and FSIM

摘要

In order to reduce the computation cost of extracting feature, a new algorithm for image retrieval is proposed. This method utilize multiple-feature including ellipse measuring as shape feature, phase congruent and gradient magnitude as the contrast feature for image retrieval in this paper. It not only can fill up a deficiency of the contrast invariant of phase congruent, but also can fully utilize scale and rotation invariance of ellipse measuring. To evaluate the effectiveness of the proposed method, we carried out a series of experiments on shape216 and three color image databases. The experiments show that the proposed method in this paper is more efficient than conventional algorithms.