A Study of Station Platform Security Using Digital Image Processing Technology

The rail transport systems, including railway, MRT, and latest high speed rail, all play a major role in Taiwan public transit circuits. According to statistics, over 2,360,000 passengers use these railway systems every day. Therefore, how to strengthen the safety and security of station platform has become an important issue in recent years. The traditional way to perform station platform security is by the use of monitoring equipment, e.g., CCD and DVR, to record the passengers' behavior on the platform. The main drawback of such a system is that it just could record the video frames and can not provide warning message to passenger immediately. Hence, this paper proposed a novel method that could detect the passenger moving direction without using additional passenger detection equipment. First, the proposed method extracts the motion vectors from the digital compressed video files, e.g., MPEG-4 or H.264, and then uses these motion vectors to determine whether passenger is approaching railway. Once passenger is too approaching railway, the proposed method will notify the passenger via broadcasting or speech message to increase the safety on the platform. Experimental results show that the proposed method can effectively identify the motion direction of passenger on the platforms, and can successfully start the broadcasting warning system when passenger is too approaching railway.