文章基本信息

标题：Devanagari Isolated Character Recognition by using Statistical features ( Foreground Pixels Distribution, Zone Density and Background Directional Distribution feature and SVM Classifier)
本地全文：下载
作者：Mahesh Jangid
期刊名称：International Journal on Computer Science and Engineering
印刷版ISSN：2229-5631
电子版ISSN：0975-3397
出版年度：2011
卷号：3
期号：6
页码：2400-2407
出版社：Engg Journals Publications
摘要：In this paper, we present a methodology for off-line Isolated handwritten Devanagari character recognition. The proposed methodology relies on a three feature extraction techniques. The first technique is based on recursive subdivisions of the character image so that the resulting sub-images at each iteration have balanced (approximately equal) numbers of foreground pixels, as far as this is possible. Second technique is based on the zone density of the pixel and third is based on the directional distribution of neighboring background pixels to foreground pixels. The 314 sized feature vector is form from the three feature extraction techniques for a handwritten Devanagari character. The dataset (12240 samples) of handwritten Devanagari Character, have been prepared by writing the different � 2 people who belongs to different age group and obtained the 94.89 % recognition accuracy.
关键词：Davanagari Character Recognition; Forground pixel; Zone density; Background directional distribution; Support Vector Machine