期刊名称：International Journal of Computer Science and Information Technologies
电子版ISSN：0975-9646
出版年度：2017
卷号：8
期号：1
页码：100-104
出版社：TechScience Publications
摘要：Content-based image retrieval (CBIR) is theapplication of computer vision techniques to the imageretrieval downside, that is, the matter of finding out digitalimages in massive databases. CBIR applies to techniques forretrieving similar images from image databases based onautomated feature extraction methods. In this paper we’veanalyzed different techniques of content-based imageretrieval. The development of CBIR systems involves analysison databases, image processing and handling problems thatvary from storage issues to friendly user interfaces. Theaccuracy of retrieval has subsequently increased with the useof low level features such as color, texture, shape etc. Frommany research activities performed on CBIR system showsthat low-level image features cannot always describe highlevelsemantic concepts within the users mind because, usinglow level features only does not include human perception. Ifhuman perception is allowed in the image retrieval system theefficiency can be improved.
