基于交叉分块直方图的图像检索方法  

An Image Retrieval Method Based on Crossed-blocking Histogram

作　　者：王蓉[1] 李冲[1] 杨宁[1] 

机构地区：[1]中国人民公安大学警务信息工程学院,北京100038

出　　处：《科学技术与工程》2014年第26期111-115,共5页Science Technology and Engineering

基　　金：国家自然科学基金(71203229);北京高等学校青年英才计划项目(YETP1366);2014北京市支持中央高校共建项目公安视听专业建设;中央高校基本科研业务费专项基金(2014JKF01121;2014JKF01122)资助

摘　　要：在基于内容的图像检索技术中,颜色是常用的图像特征。提出了一种基于交叉分块直方图的图像检索方法。首先,对图像进行交叉分块得到若干子图像;然后提取子图像的颜色直方图特征,构造特征向量;最后根据空间距离度量图像间的相似程度,确定检索输出结果。用交叉方式分块得到的子图像间相互交叉重叠,能够在获取图像颜色特征空间分布信息的同时保持图像目标的完整性,并有效抑制图像背景区域对检索结果的干扰。仿真实验结果表明,基于交叉分块直方图的图像检索方法具有优于基于均匀分块直方图的图像检索方法的性能,适用于对多种类型图像的检索。Color is the common feature of the Content-based hnage Retrieval. An image retrieval method basedon crossed-blocking histogram is proposed. Firstly, blocking the image crossly to obtain several sub-images; sec-ondly, extracting the color histograms of the sub-images, which compose the eigenvector of the image; thirdly,measuring the similarities between different images through the spatial distance, and the output images are then de-termined by the similarities. These sub-images blocked from one image are crossed and overlap and contain the spa-tial color feature, which can maintain the integrity of the image target and restrain the disturbance of the retrievalresults brought by image background effectively. According to the simulating experiments, the image retrieval meth-od based on crossed-blocking histogram can get the better performance than the method based on uniform-blockinghistogram and retrieve multiple types of images.

关 键 词：图像检索 颜色直方图交叉分块 相似性度量 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]