Font Size: a A A

The Research On Key Techniques Of Content-based Image Retrieval (CBIR)

Posted on:2009-06-09Degree:MasterType:Thesis
Country:ChinaCandidate:Q XuFull Text:PDF
GTID:2178360242478139Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
With the rapid development of computer, Internet techniques as well as the application of multimedia leading to the number of images growing at a remarkable rate, how to organize, manage and retrieval large image database is becoming a very important and challenging subject. Content-based image retrieval (CBIR) becomes one of the most active research focuses of implementation of multimedia.In this paper, according to the hot point of the current research of CBIR, based on the analyzing and discussing for the key techniques of CBIR, we mainly research the image retrieval algorithms based on overall and local characteristics .The main content of this paper are summarized as follows:1. Some key techniques and algorithms of CBIR are deeply analyzed and discussed. Moreover, we compare some classical methods in the same testing environment.2. A novel algorithm for image retrieval based on edge points is presented. It takes not only edge angle distribution structure into consideration, but also the space color information of edge points. Experimental results show that this algorithm is simple and improves the image retrieval efficiency.3. A novel approach using combined features to retrieve images is presented. The feature detection in this work is an integrated process: edges and interest points are detected directly based on the Harris function , so complex of algorithm is reduced. Experimental results show this system has good performance.4. We design a CBIR system as the test platform for various image retrieval algorithms, which is an experimental frame system. The operation system platform for developing our CBIR test system is Windows XP, and the development environment is Matlab7.1.
Keywords/Search Tags:Content-based image retrieval, Region, Interest points, Annular histogram based on interest points, The relevance feedback
PDF Full Text Request
Related items