期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：2
页码：1340
DOI：10.15680/IJIRCCE.2017.0502007
出版社：S&S Publications
摘要：This research proposes an efficient algorithm based on SURF (Speedup Robust features), SVMclassifier (Support Vector Machine) classifier and Bayesian classifier for Content Based Image Retrieval. CBIR ischallenging task which retrieves the similar images from the large database. Performing feature of image matching, thefalse matching points are eliminated through this algorithm. From rest of the match point, we estimate the spacegeometric transformation parameters between two images and thus matching process is completes. The SURFalgorithm is used to detect and descript the interest points and match them and then fed them into SVM and BayesianClassifier for further classification.
关键词：Content Based Image Retrieval; SURF algorithm; Support Vector Machines and Bayesian
