History-Based Dynamic Hand Gesture Recognition
HUANG Guo-fan,CHENG Xiao-ping School of Computer and Information Science,Southwest University,Chongqing 400715,China
A method of hand gesture recognition is proposed using appearance-based modeling in this paper. Hand gesture is represented by history-based approach,building a hand model using skin history image. Hand gesture is recognized by prototype template matching. Tangent distance is used for the first time to measure the similarity between the test sample and the template,thus eliminating visual sensitivity. The experiment showed that our method has markedly improved recognition rate.
【CateGory Index】： TP391.41