文章基本信息

标题：Offline Handwritten Word Recognition using Multiple Features with SVM Classifier for Holistic Approach
本地全文：下载
作者：Shruthi A ; M S Patel
期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2015
卷号：3
期号：6
出版社：S&S Publications
摘要：Offline Handwritten Word Recognition (HWR) is an interesting research field in the domain of digitalimage processing. Offline Handwriting recognition has achieved a high attention for many years due to its keycontribution in the digital libraries extension. Offline recognition of handwritten words by multiple and differentwriters is still an exposed problem, because of multiple challenges including strong variability in writing style and eachperson have their own control over writing. English Handwritten words recognition is most important now daysbecause English script is widely used language in the world and most of the countries used as official language. Toidentify the handwritten words, each word take is an individual entity so holistic approach is used. In the proposedmethod reports multiple features namely: density features, long run features and structural features for extraction in theinput handwritten document image. Next phase is classification and this phase is most important for word recognition,for classification Support Vector Machines (SVM) classifier is used. To estimate the performance, own dataset iscreated and sample words for testing collected from multiple people. There are so many applications in HWR like,postal address identification, historical document conversion, signature verification and etc. This presented work gives88.13% of recognition rate.
关键词：Handwritten Word Recognition (HWR); dataset; density features; long run features; structural features;classifiers.