期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2015
卷号：4
期号：8
页码：7469
DOI：10.15680/IJIRSET.2015.0408142
出版社：S&S Publications
摘要：Now a days shape recognition has got more importance because of its wide range of application and use.Each and every day researchers are coming out with their new innovative ideas and executions, but still it is a openproblem for the researchers. In the proposed method, 2D shape characterization technique is explained using ContentBased Image Retrieval (CBIR) system. CBIR is the effective method towards identifying the similar shaped imagesbased on the 3 factors. First one is morphological filters which make use of operations like erosion, dilation, openingand closing to find the similar images. In the next method contours of the images playa a major role in identifying thematch. Third method is the combination one, which combines the above two methods to find the similar images fromthe large database. Support Vector Machines (SVM) classifier is used to find the accuracy of the matched objects.Experimental results show the good accuracy rate.
