文章基本信息

标题：Retrieval Of Digital Images Based On Shape Feature Using Support Vector Machines And Self Organizing Maps
本地全文：下载
作者：Sarita Sharma ; Avinash Dhole
期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2014
卷号：3
期号：4
页码：1273-1279
出版社：Shri Pannalal Research Institute of Technolgy
摘要：Legendre moments are continuous moments, hence, once applied to discrete-space images, numerical approximation is concerned and error happens. This paper proposes a technique to work out the precise values of the moments by mathematically desegregating the Legendre polynomials over the corresponding intervals of the image pixels. Experimental results show that the values obtained match those calculated in theory, and therefore the image reconstructed from these moments have lower error than that of the standard ways for constant order. Though constant set of tangible Legendre moments are often obtained indirectly from the set of geometric moments. Content Based Image Retrieval is the application of computer techniques to resolve the matter of checking out digital image within the giant information. In Content Based Image Retrieval, images are retrieved based on color, texture and shape. The CBIR system uses these options for retrieval of images and therefore the technique for obtaining these options is understood as Feature Extraction. For image classification I actually have worked on the feature Shape. Content Based mostly Image Retrieval system exploiting Accurate Legendre Moments (ALM) for grey scale images and color images is projected in my work. Further, the image classification potency is improved by using Support Vector Machine (SVM) and Self organizing Maps (SOM) classifiers.
关键词：Content Based Image Retrieval; Legendre ; Moments; Accurate Legendre Moments ;Active Contour ; Without Edges; Shape; Support Vector Machines ;Self ; Organising Feature Maps.