Journal Paper

Paper Title - Contourlet Transform Based Feature Extraction For Handwritten Malayalam Character Recognition Using Neural Network


Abstract
Optical Character Recognition (OCR) is one of the important fields in image processing and pattern recognition domain used to recognize printed and handwritten characters. Handwritten character recognition has always been a challenging task due to its substantial variation in appearance. This paper present an efficient and robust algorithm for recognition of handwritten isolated Malayalam character. The proposed system consists of image acquisition, preprocessing, segmentation, feature extraction, classification & recognition stages. Because of the curved nature and no inherent symmetry of Malayalam characters, its feature extraction is difficult. So the main aim of this paper is to propose a fast and easy to use, feature extraction method that gives a good performance for Malayalam character recognition .Contourlet transform is used for feature extraction in addition to ratios of grid values in horizontal and vertical directions. A feed forward artificial neural network trained using the back propagation algorithm is being used as the classifier. The proposed system achieves a maximum recognition accuracy of 97.3 %


Author - Aji George, Faibin Gafoor

Citation - Aji George   ,   Faibin Gafoor   ,   Aji George, Faibin Gafoor " Contourlet Transform Based Feature Extraction For Handwritten Malayalam Character Recognition Using Neural Network " , International Journal of Industrial Electronics and Electrical Engineering , Volume-2,Issue-4  ( Apr, 2014 )

Indexed - Google Scholar


| PDF |
Viewed - 1
| Published on 2014-06-19