Target Detection And Alarm Research Of The Railcar Video Surveillance System

During the railcar running, the object left behind in orbit on the line, intrusion gauge locomotives moving target will affect the safety of running and railcar operations. Currently railcar video surveillance system can be capable of monitoring the process of railcars to run and shoot video images, but usually used to assist the driver to operate, and provide a view of video data afterwards. Since there is no complete suspicious intrusion detection system for novelty or moving targets, so it has a lot of passivity on the monitor, and can’t fully play its role to protect the safety of the railcar.In this paper, according to the use of railcars, operating characteristics, and job safety analysis, proposed the introduction of target detection based on video images getting from the railcar video surveillance systems and using background subtraction method by analyzing the advantage to complete the detection of the novelty on the rail and moving. For different situations, we use the method based on color features and texture features with creating a background image model and detecting novelty and moving targets using the method of background subtraction between the background image and foreground image.In the end, we propose to combine the railcar video surveillance system with GYK and a systematic program on how to build the system and how to achieve communication between the railcar video surveillance and GYK and combination control.