基于颜色-空间的图像检索算法

Algorithm of Image Retrieval Based on Color-Space

在分析了现有的基于内容的图像检索方法的基础上,提出了一种基于颜色空间分布的图像检索方法。该方法采用HSI颜色模型提取代表色,提高了图像检索的速度;通过图像分块后统计各方块颜色的代表色直方图并赋予不同的权值,提高了图像检索的精度,改善了图像检索的质量。实验结果表明,排序上明显优于其它几种分块算法,可以得到满意的查询结果。

This paper deals with a new image retrieval method which can improve the retrieving speed by extracting representative color in HIS color pattern and enhance the retrieval precision and quality by providing different parameter of histogram for each segment. Experiments show that ideal retrieval results can be obtained with this method.