Chinese Cursive Script Character Image Retrieval Based on an Integrated Probability Function

Often in content-based image retrieval, a single image attribute may not have enough discriminative information for retrieval. On the other hand, when multiple features are used, it is hard to determine the suitable weighting factors for various features for optimal retrieval. In this paper, we present an idea of integrated probability function and use it to combine features for Chinese cursive script character image retrieval. A database of 1400 monochromatic images is used. Experimental results show that the proposed system based on Legendre moment feature, Zernike moment feature, and pseudo Zernike moment feature is robust to retrieval deformed images. Using our integrated probability function, ninety-nine percent of the targets are ranked at the top 2 positions.