Simulation Analysis of Embedded Visual Monitoring Method for Automobile's Abnormal Movement
Xiaofen Wang · Jisuanji fangzhen · 2014
In this paper,accurate visual monitoring problem of automobile's abnormal movement was studied.Abnormal movement of the automobile refers to the characteristics which are different from the normal driving behaviors. The similarity in automobile's movement process will bring greater interference to automobile's abnormal movement behavior identification. To solve these problems,this paper presented a method for visual monitoring of automobile's abnormal movement based on embedded technology combining with optical flow motion recovery algorithm. An embedded monitoring procedure was designed to make optical flow motion recovery for the automobile's visual monitoring image acquired by sensor,which can fight against the behavior interference caused by slight movement and shake of automobile and provide the basis for visual monitoring of automobile's abnormal movement. The results of optical flow recovery for automobile's monitoring image are made background separation to get the monitoring results of automobile's abnormal movement. Experimental results show that the improved algorithm for embedded visual monitoring of automobile's abnormal movement,can eliminate the interference,greatly improve the accuracy of monitoring and complete the accurate identification of automobile's abnormal movement.