基于DCT和SVD的图像检索算法  

Image Retrieval Algorithm Based on DCT and SVD

作　　者：许相莉[1] 张利彪[1] 于哲舟[1] 周春光[1] 

机构地区：[1]吉林大学计算机科学与技术学院,长春130012

出　　处：《吉林大学学报（理学版）》2008年第6期1125-1130,共6页Journal of Jilin University:Science Edition

基　　金：国家自然科学基金重大项目基金(批准号:60433020;60673099;60773095);国家高技术研究发展计划项目基金(批准号:2007AA04Z114);欧盟项目TH/AsiaLink/010(批准号:111084);教育部"符号计算与知识工程"重点实验室项目基金(批准号:02090);吉林省科技发展计划项目基金(批准号:20080168);吉林大学"985工程"项目基金(批准号:20080236)

摘　　要：提出一种在子块分割和区域划分的基础上,利用离散余弦变换和奇异值分解对图像进行特征提取的检索算法.首先对图像进行子块分割,利用离散余弦变换提取重要系数作为子块颜色特征,进而对图像进行区域划分,将每个区域中的子块颜色特征分量组成矩阵进行奇异值分解,得到该区域的检索特征向量,从而完成图像检索.实验结果表明,该算法取得了较好的查全率和查准率,具有较好的检索效果.A feature extraction and retrieval algorithm based on child-block partition and area division using discrete cosine transformation and singular value decomposition is proposed. First the image is divided into child-blocks, discrete cosine transformation is used to extract important coefficients as color feature of the child-block, and then the image is divided into areas, some matrixes are constructed of the discrete cosine transformation coefficients in this area, which are submitted to singular value decomposition to gain the retrieval feature vector to accomplish image retrieval. Experiments show that the proposed algorithm gains good recall and precision, and has a good retrieval effect.

关 键 词：离散余弦变换 奇异值分解 图像检索 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]