A new texture feature extraction method for image retrieval

Contrast, as a type of Tamura texture feature, is a global variable which can well describe the statistical distribution of the brightness in the entire image, but can not reflect the local brightness information of the image. This paper proposes an improved method, which makes use of the statistical moments of intensity histogram to extract more information from the image. Tested on a tire tread pattern dataset, the proposed method provides better retrieval performance compared with existing methods. Hence, it is concluded that the proposed method is effective in texture feature description especially for type texture, due to the fact that more intensity information is exploited.