期刊名称：International Journal of Engineering and Computer Science
印刷版ISSN：2319-7242
出版年度：2014
卷号：3
期号：12
页码：9755-9760
出版社：IJECS
摘要：Content-based image retrieval utilizes representations of features that are automatically extractedfrom the images themselves. Allmost all of the current CBIR systems allow for querying-by-example, atechnique wherein an image (or part of an image) is selected by the user as the query. The system extractsthe feature of the query image, searches the database for images with similar features, and exhibits relevantimages to the user in order of similarity to the query. In this context, content includes among other features,perceptual properties such as texture, color, shape, and spatial relationships. Many CBIR systems have beendeveloped that compare, analyze and retrieve images based on one or more of these features. Some systemshave achieved various degrees of success by combining both content-based and text-based retrieval. In allcases, however, there has been no definitive conclusion as to what features provide the best retrieval. In thispaper we present a modified SVM technique to retrieve the images similar to the query image
关键词：CBIR ; SVM ; Content Based Image Retrieval; Modified SVM
