1. Optical character recognition (ocr) system for english language using artificial neural network (ann) classifier
Authors : Honey Mehta, Sanjay Singla
Pages : 42-48
DOI : http://dx.doi.org/10.21172/1.73.506
Keywords : Unicode, Artificial Neural Network, Scanned images. Abstract :Character recognition is a very complex task when one is talking about record keeping of all the data in digital format which can be further used to perform various manipulation operations. The actual problem is character recognition of various fonts in which the text is written. We proposed a new approach with the concept of Artificial Neural Network(ANN) for character recognition from scanned images. Once the data is implemented by an OCR(Optical Character Recognition) system and converted into digital form, it can be edited, manipulated and stored in computer understandable form. Performance of our system is observed by considering the variations in number of iterations and the variations in the number of characters. The proposed system produce 98.89 % recognition rate for two different fonts( latin arial and latin tahoma) by considering up to 90 characters at a time.
Citing this Journal Article :Honey Mehta, Sanjay Singla, "Optical character recognition (ocr) system for english language using artificial neural network (ann) classifier", Volume 7 Issue 3 - September 2016, 42-48
Click here to Submit Copyright Takedown Notice for this article.