Abstract: Abstract: Content based image retrieval is a research focus in clothing electronic commerce. Shape feature is the most important clothing image visual features. Image retrieval based on shape feature has been the problem of image processing for many years. In order to improve the accuracy of the system, an algorithm has been put forward which use general Fourier descriptors and the improved feedback technology .General Fourier descriptor has outstanding performance in all kinds Fourier descriptors. At the same time, according to the probability distribution, the algorithm improves weight calculation method for the query vector .So the correlation dimension is more reasonable. The experimental results show that the proposed algorithm outperforms the traditional algorithm.

Key words: image retrieval, clothing, Fourier descriptor, relevance feedback