The Study Of Gesture Recognition Based On Video

The vision based gesture recognition technology is a hot research topic. The theory and the background of the vision based gesture recognition are summarized. This paper proposes a novel motion detection algorithm based on the combination of interframe difference and skin color. The paper extracted the boundary chain code and barycenter of area firstly. Then it designs a gesture recognition system based on Iterated Conditional Modes (ICM). At last, we validate the algorithm through experiments and analyze its performance.This work includes the following aspects:1、A novel motion detection algorithm is proposed. The advantage of the algorithm is ability of getting rid of minor information in low resolution.2、A gesture recognition system based on Iterated Conditional Modes (ICM) is designed. The system uses the gesture start frame and end frame step-by-step identification method to improve the recognition performance and save so much time.3、Experimental results show the effectiveness of discrete ICM-based gesture recognition algorithm. We classify the extracted signemes as correct, partially correct or incorrect depending on whether both the start and end locations are correct, only one of them is correct or both are incorrect, respectively. Out of the136extracted video signemes,119were correct,17were incorrect.