Real-time Dynamic Hand Gesture Recognition using Hidden Markov Models

The goal of interaction between human andcomputer is to find a way to treat it like human-humaninteraction. Gestures play an important role in human’s daily lifein order to transfer data and human emotions. The gestures areresults of part of body movement in which hand movement is themost widely used one that is known as dynamic hand gesture. So itis very important to follow and recognize hand motion to providemulti-purpose use. In this paper, we propose a system thatrecognizes hand gestures from continuous hand motion forEnglish numbers from 0 to 9 in real-time, based on HiddenMarkov Models (HMMs). There are two kinds of gestures, keygestures and link gestures. The link gestures are used to separatethe key gestures from other hand motion trajectories (gesturepath) that are called spotting. This type of spotting is a heuristicbasedmethod that identifies start and end points of the keygestures. Then gesture path between these two points are given toHMMs for classification. Experimental results show that theproposed system can successfully recognize the key gestures withrecognition rate of 93.84% and work in complex situations verywell.

Keywords:

Hand tracking , Gesture path , key points , Dynamic hand gesture recognition , Hidden Markov Model