本系统通过将基于改进的加权的局域颜色直方图的图像检索方法和全局直方图的图像检索方法相结合，提高查全率和查准率。其中，基于分块局域直方图的检索方法利用了图像中间部分的重要性，将图像平均划分成3×3个子块，取中间一块的图像，计算其与参考位图相应位置的颜色特征距离，再计算原图的颜色直方图与参考位图的颜色特征距离，分别赋予权值后得出的值就是图像之间内容的相似程度。本文引入欧氏距离的相似性度量方法实现图像检索。实验表明，该方法具有较好的查全率和查准率。-The system through a combination of improved weighted local color histogram-based image retrieval method based and global image retrieval method to improve the recall rate and precision rate. Wherein chunked local histogram-based retrieval method using the importance of the intermediate portion of the image, the image is evenly divided into 33 sub-blocks, take the middle piece of the image color characteristic distance calculated with reference to the bitmap corresponding positions , then calculate the original color histogram and color characteristics of the reference bitmap distance, respectively assigning weights to the value obtained is the degree of similarity between the images. This paper introduces the Euclidean distance similarity measure image retrieval. Experiments show that the method has good recall and precision.