综合颜色特征的图像检索方法研究

Research on Image Retrieval Method with Comprehensive Color Feature

摘要: 为了克服颜色矩对图像颜色特征描述不够准确，且缺少空间特征信息而导致图像检索查准率低这一缺陷，本文提出了一种综合图像颜色特征检索方法，将图像予以分类并考虑空间信息，利用颜色矩、图像分块、熵和能量分类的方法进行检索。通过计算机模拟和实验，结果表明该方法较显著提高检索结果的准确性。

Abstract: In order to overcome the defects of color moment whose description of color feature is not accurate and lack of spatial information that causes a low retrieval precision, this paper presents a new image retrieval method with comprehensive color feature. The method classifies images and takes their spatial information, then retrieves image by using color moment, image block, image entropy and energy. The result of computer simulation and experiment shows that the method noticeably improves retrieval precision.