Research Of Traffic Police Gesture Recognition Algorithm Based On Digital Image Processing

With the continuous development of society, people's living standard has been increasingly improved, so did the quality of life. Cars has become one of the important transportations, this will give road an unprecedented pressure. As time goes on, the pressure will be more sand more heavy. The way to control road traffic need further diversity and intelligent. Traffic police gesture recognition is appeared to solve this problem, and has achieved great development. Gestures have the advantages of flexibility and diversity. And the analysis of gestures also referred to the processing of background noise, so the gesture recognition is a difficult subject.This paper involves gesture recognition in all its aspects. In a previous study basis, it has further summarized and innovated, forming a static gesture recognition system based on vision. This system can be divided into three parts:the image preprocessing part, the boundary feature extraction part and the part of gesture recognition. Image preprocessing can eliminate the noise, to obtain a clear target image. The part of boundary feature extraction use a chain code to describe the boundary of the target, and then normalize it, at last it can get the shape several. The last part of the system can realize the comparison of feature vector, and by the judgment the Euclidean distance to display similar degree among the images. Finally, it can judge the specific meaning of the gestures. This method can realize the recognition of the gesture of traffic, and has a high accuracy.