A Novel Method of Tamil Character Recognition Using Slope Method
S. Selvakumar Raja, Mala John · SSRN Electronic Journal · 2010
Tamil, a south Indian language, is one of the oldest languages in the world. This paper presents an efficient slope method for recognizing Tamil characters based on extracted features like horizontal lines, vertical lines, and curves. Various preprocessing operations like segmentation, thresholding, skeletonization (or) thinning and normalization were performed on the digital image to enhance the quality of the image. The extracted features were used to train and test the neural network. The proposed system has achieved very good recognition rate on the Tamil character database. The results obtained are comparable to the best ones reported in the recent literature.