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：2016
卷号：4
期号：2
页码：1540
DOI：10.15680/IJIRCCE.2016.0402129
出版社：S&S Publications
摘要：The paper construe a technique for medical images annotation based on the SURF descriptor, cross validation and color histogram. The CBIR is a substantial research field for maneuver a large extent of image databases. This paper concentrates on a decisive algorithm which is based on Speeded up Robust Features, Cross Validation and Color Histogram. Firstly in this method the SURF featuredetector is applied for the extraction of reference images and then performs matching of feature points present in the image, respectively. Finally, the estimation of the space geometric transformation parameters between two images is performed according to the rest of the match point and this will conclude the process of matching. The interest points are used for the extraction of the similar images but with different pose and certainty. Here, the same thing is done to retrieve this but with the help of SURF, Color Histogram and the obtained data is fed to the Cross Validation for performing further classification. The implementation of this proposed work is done with the help of the Image Processing Toolbox under Mat lab
