A Relevance Feedback Retrieval Method Based on Tamura Texture

This paper presents a relevance feedback method to be working out well for trademark retrieval. For the semantic gap between the low-level similarity and the high-level user´s query in content-based image retrieval, this paper proposes a retrieval strategy to remedy the semantic gap. One side is to use the Tamura texture features which consistent with human vision perception and the low-level feature of images. On the other side use support vector machines to train an optimal margin hyper-plane for classification. Then based on the results we moderate the feature to further retrieval. Experimental results show that the method has good effectiveness for moderate scale trademark database.