A Novel Bilingual OCR System Based on Column-Stochastic Features and SVM Classifier for the Specially Enabled

All sections of the society need to benefit from the strides in Information Technology, more so the differently enabled. This paper presents a Novel solution for a text read out OCR system adapted for the visually challenged. A paper of text from Malayalam magazines, newspapers, books or journals placed on a flatbed scanner would be recognized and read out as text. Selected text could also be printed with a Braille Embosser. The paper addresses all issues related to the integration of such a system. This novel system helps to bring reading rooms and libraries to the specially enabled populace. It finds application in all areas where documents need to be accessed. This system designed has an inherent advantage in the sense that it can also be of functional utility to the literate and semi-literate citizen, besides to visually affected persons. Such systems are not presently available for Indian languages. This system has been tested and evaluated on Malayalam scripts.