Study on HMM_SVM-based Chinese Sign Language Recognition
Weining Zhao · Journal of Changchun University · 2011
This paper presents a Hidden Markov-based Chinese sign language recognition method by means of image capture of gestures,image processing and dimensionality reduction techniques.Through the application of Sugeno fuzzy integral,image processing technology is basic step for computer graphics identification,which can separate the hand-shaped area from the background by using histograms features,and then removes the part outside hand from obtained hand image by dimension reduction process to get the profile of hand.This method realizes the static identification for simple sign language without related gloves tools,the correct rate reaches 85% or more and the feasibility of sign language recognition in China is proved.