ARABIC SIGN LANGUAGE RECOGNITION SYSTEM ON SMARTPHONE

Deaf And other verbally challenged people face challenges most of the time communicating with the society, sign language is what they commonly use between them to represent what they want to say to each other for example numbers, words or phrase. To bridge this communication Barrier between them and the society an automated system to stand as a translator between them and the society is needed, recently many researches are performed, but most of the developed Systems are only executable on computers, which are difficult and impractical to take around. We are proposing the use of smartphone as a platform, a client server system is implemented, sign image background is detected and removed under HSV color space, features extracted from the frame are the binary pixels, SVM is used to classify the features, we are able to classify 10 Arabic Sign Language with an experimental accuracy result of 92.5%.