Research Of The Subway Security Monitoring System Based On Facial Features Analysis Technology

In recent years, intelligent video analysis technology has gradually become a hot topic in computer vision, artificial intelligence and security monitoring. However,previous research mostly focused on middle and low level technology, and the research on new situations or high level technology was not enough, so it’s difficult to adapt to the needs of social development. Therefore, more research on the intelligent video analysis technology at the abstraction semantic layer is imperative.To improve the prediction accuracy of the security and protection system and achieve multi-target behavior analysis in surveillance scene, this paper presents a method of monitoring the subway security based on facial features analysis technology. Facial features analysis includes static facial expressions recognition,assessing the degree of confusion of the wrinkle in the glabellar area, similarity comparison of left and right mouth, face type recognition, the changes of the pupil position.etc. The initial stage of the software system is mainly to complete the face detection under complex background. Then preprocessing images obtained,localizing the facial features and perform other operations. Finally, designing and testing each module in the Visual Studio 2010 and matlab 8.0. The joint commissioning results showed that this system has strong real-time processing capability and high stability.