Tamil Sign Language to Speech Translation

摘要

The Sign Language Recognition is the most popular research area involving computer vision, pattern recognition and image processing. Many techniques have been developed recently in these fields. Sign languages are used for communication and interface. There are various types of systems available for Sign Language Recognition. Sign Language Recognition is a very challenging research area. In this paper, a system to recognize static gestures representing Tamil words and speech translation has been proposed. The approach used in this paper is robust and efficient for static hand gesture recognition. Sign language has different applications in many domains like HCI (Human Computer Interaction), Robot control, Security, Gaming, Computer vision etc. A different approach is handled in this work for recognizing Tamil Sign Language. It consists of three phases. The first phase is preprocessing where the images are processed through the steps like resizing, gray conversion, filtering for reducing the distortion and black and white conversion. Second phase is feature extraction. Shape descriptors such as solidity, eccentricity, perimeter, convex area, Major axis length, Minor axis length and orientation are applied to the black and white images to extract the features. Third phase is the classification where a Naive Bayesian classifier is used to recognize the signs from trained set of gestures. The features derived are used to train the classifier first and then the testing images have been introduced for classification. The proposed system is able to recognize images with 90% accuracy.