Real Time Hand Gesture Recognition Using Sift

The objective of the gesture recognition is to identify and distinguish the human gestures and utilizes these identified gestures for applications in specific domain. In this paper we propose a new approach to build a real time system to identify the standard gesture given by American Sign Language, or ASL, the dominant sign language of Deaf Americans, including deaf communities in the United States, in the English-speaking parts of Canada, and in some regions of Mexico. We propose a new method of improvised scale invariant feature transform (SIFT) and use the same to extract the features. The objective of the paper is to decode a gesture video into the appropriate alphabets.