Recognition and Classification of Broken Characters using Feed Forward Neural Network to Enhance an OCR Solution
Journal Title: International Journal of Advanced Research in Computer Engineering & Technology(IJARCET) - Year 2012, Vol 1, Issue 8
Abstract
The major problem faced by an Optical Character Recognizer (OCR) can be attributed to the presence of a large number of broken characters in a digital document image. Recognition of such documents accurately, that contain abundant broken characters still remains a challenge to most of the OCR solutions. Multi Layer Feed Forward Neural Network Classifier (MLFNC) can be used to enhance the efficiency of an OCR. MLFNC tries to improvise the recognition by classifying broken characters into a different group. This gives an opportunity to process broken characters in a more effective manner separately. Here, a new method has been proposed which uses feed forward neural network to classify broken characters prior any processing is done by an OCR with a considerable accuracy. MLFNC is a simple network with a very small time complexity due to which, there is a least effect on the time complexity of the solution provided by OCR.
Authors and Affiliations
Manas Yetirajam, , Manas Ranjan Nayak, , Subhagata Chattopadhyay
A Combined Approach to Part-of-Speech Tagging Using Features Extraction and Hidden Markov Model
Words are characterized by its features. In an inflectional language, category of a word can be express by its tense, aspect and modality (TAM). Extracting features from an inflected word, one can categorised it wi...
CONTENT BASED IMAGE RETRIVAL USING FRACTAL SIGNATURE ANALYSIS
Content-based image retrieval has become a reliable tool for digital image applications, especially for image retrieval purposes. There are several advantages of the image retrieval techniques compared to other simple re...
ENHANCED TECHNIQUE FOR SECURED AND RELIABLE WATERMARKING USING MFHWT
A general watermarking techniques are used for copy right protection. In this watermarking scheme should achieve the features of robustness and imperceptibility. This paper represents the watermarking algorithm in the DW...
Earthquake Disaster based Efficient Resource Utilization Technique in IaaS Cloud
An Automated Mobile Applications Generator for Remote Database Access
Due to the display area and resource restrictions of small mobile devices, currently, most reported database query systems developed for them are only offering a small set of pre-determined queries that can be po...