期刊名称：International Journal of Computer Science & Technology
印刷版ISSN：2229-4333
电子版ISSN：0976-8491
出版年度：2013
卷号：4
期号：2
页码：579-584
语种：English
出版社：Ayushmaan Technologies
摘要：Optical character recognition is used widely for generating digital computer data from printed or handwritten text document. Devanagari script is basis for many Indian languages. A lot of work has been completed in optical character recognition of Devanagari script. Devanagari script consists of numerous fundamental characters like partial consonants, full consonants or character, vowel modifiers, and diacritics. Segmentation of touching or fused characters of Devanagari script is sometimes difficult due to interline space or overlapping and noise. The algorithm used here is for segmentation of fused Devanagari characters into its constituent partial or full consonants. The structural properties of Devanagari script is used with the consonant or conjunct characters height and width for finding the best point of segmentation of fused characters.
关键词：Devanagari Script;Conjunct Character;Optical Character Recognition;Segmentation
