An automated vision based recognition system for Sri Lankan Tamil sign language finger spelling

As computers become more and more pervasive in human lives, the need for natural and effective Human Computer Interaction (HCI) becomes more important than ever. Speech recognising and voice commanding remain to play an important role in the HCI field. However, these systems are restricted for deaf community. In Sri Lanka, the native language of the deaf community is Sri Lankan sign language, which defines set of vocabulary of gestures corresponding to frequently used words. If a word is not defined, they are spelt out the word using gestures that correspond to the letters in the Sinhala or Tamil alphabet. In this paper, authors investigate as regards Sri Lankan Tamil sign language finger spelling alphabet, problem of recognising Sri Lankan Tamil finger spelling from vision based recognition and technical challenges behind it.