A New Content-Based Image Retrieval

Color features are important to pictures and they are easy to calculate. Therefore, the features are widely used in content-based image retrieval (CBIR). A color histogram describes the frequency of colors in images. It won't change with the variations of pictures' geometry, so it is a widely used feature for image indexing although it has some shortcomings. In this paper, we propose a new color space quantization arithmetic and a new histogram similarity function with respect to the shortcoming and make a prototype system for image indexing. The experiments show that the new color image retrieval algorithm is effective and robust