基于模糊颜色量化的图像检索*

颜色量化是基于颜色特征的图像检索的一个重要方法。在颜色量化中引入模糊集合理论，提出了一种基于人的主观视觉感知的模糊颜色量化方法，以减小量化误差和适应人的感知模糊性，根据此方法提取模糊颜色-空间特征用于图像检索。实验结果显示了模糊量化方法的有效性和检索算法较高的检索准确性。

Image Retrieval Based on Fuzzy Color Quantization

Color quantization is an important method of image retrieval based on color feature. By introduction of the fuzzy set theory, a fuzzy color quantization method based on the subjective perception of human is proposed to minimize color quantization error and adapt to the fuzziness of human's perception. Based on the method, fuzzy color-spatial feature is extracted for image retrieval. Experimental results show effectiveness of the quantization method and significant high retrieval accuracy of the algorithm.