INTERNATIONAL JOURNAL OF NOVEL RESEARCH AND DEVELOPMENT International Peer Reviewed & Refereed Journals, Open Access Journal ISSN Approved Journal No: 2456-4184 | Impact factor: 8.76 | ESTD Year: 2016
Scholarly open access journals, Peer-reviewed, and Refereed Journals, Impact factor 8.76 (Calculate by google scholar and Semantic Scholar | AI-Powered Research Tool) , Multidisciplinary, Monthly, Indexing in all major database & Metadata, Citation Generator, Digital Object Identifier(DOI)
With over 300 million speakers, Hindi is India's most spoken language. The Optical Character Recognition (OCR) systems designed for the Hindi language have a very low recognition rate due to the lack of character separation in Hindi texts compared to English texts. An Artificial Neural Network (ANN)-based OCR for printed Hindi text written in Devanagari script is proposed in this paper to increase its efficiency. One of the significant purposes behind the unfortunate acknowledgment rate is blunder in character division. The fact that the scanned documents contain touching characters makes the process of segmentation even more difficult. As a result, designing an efficient method for character segmentation presents a significant challenge. A general OCR consists of preprocessing, character segmentation, feature extraction, classification, and recognition at the end.
The paper looks at the preprocessing tasks of converting grayscaled images to binary images, rectifying images, and segmenting the text of the document into paragraphs, lines, words, and then basic symbols. The neural classifier recognizes the fundamental symbols that were obtained as the fundamental unit through the segmentation process.
In this work, three element extraction procedures : histogram of projection in light of mean distance, histogram of projection in light of pixel worth, and vertical zero intersection, have been utilized to work on the pace of acknowledgment. Even distorted characters and symbols can have their features extracted using these powerful feature extraction methods. A back-propagation neural network with two hidden layers is used to build the neural classifier.
For printed Hindi texts, the classifier is trained and tested. It is possible to achieve a performance with a correct recognition rate of roughly 90%.
"Ocr for Hindi Language ", International Journal of Novel Research and Development (www.ijnrd.org), ISSN:2456-4184, Vol.8, Issue 4, page no.e406-e410, April-2023, Available :http://www.ijnrd.org/papers/IJNRD2304454.pdf
Downloads:
000118751
ISSN:
2456-4184 | IMPACT FACTOR: 8.76 Calculated By Google Scholar| ESTD YEAR: 2016
An International Scholarly Open Access Journal, Peer-Reviewed, Refereed Journal Impact Factor 8.76 Calculate by Google Scholar and Semantic Scholar | AI-Powered Research Tool, Multidisciplinary, Monthly, Multilanguage Journal Indexing in All Major Database & Metadata, Citation Generator
Facebook Twitter Instagram LinkedIn