GMT feature extraction for representation of BIM sign language

This paper focuses on the features extraction from the hand region segmentation involved in the earlier pre-processing module for Bahasa Isyarat Malaysia (BIM) sign language recognition. This Human Computer Interaction (HCI) research intention is to provide a system for aiding those with impaired hearing community to communicate with the public community in Malaysia specifically. Vision based technique has been proposed for data collection technique. Gestural Motion Trajectories (GMT) are used to keep track the movement of the hand. The objectives of this paper is to determine whether the feature extraction methods used can produced good features to get the discriminate value to distinguish each sign language.