Synchronic Detection Lines Method And Its Application In Rail Transportation

Video Detection produces a new application area in rail transportation. Generally, video Detection is prosperous in the road transportation. As a sub-system of Intelligent Transportation Systems (ITS), video detection system is employed to obtain the information of road transportation, including vehicle flow rate, average speed, queue length, lane changes, vehicle classification and license identification. Video detection plays an important role in the road transportation management. As a comparison, the research of rail video detection is rare, the application system in rail is also little known. But in the tendency that the computer is becoming the main tool in the rail transportation management, the situation of research will be changed.This paper explores the algorithms of rail video detection and proposes the method of Synchronic Detection Lines (SDL). SDL see the scene as the structural image at as the hierarchy has three layers: the detection region layer, the synchronic detection line layer and the pixel layer, then process it and obtain the interest information. The research of two aims comprised this paper. The first is detecting the direction of movement of train in crossroad and the second is obtaining the dynamic background and tracking the movement of train. In the first aim the optical flow based on the SDL is used and the second aim the probabilistic background model based on SDL is adopted. Another important content in this paper is rail detection. The rail keeps the important information about the movement of train and the SDL. This paper gives a new method to obtain the information of rail based on the Hough Transformation.The chapter four, five and six involve the design of rail video detection system. The chapter four accounts for the common contents of application system, and briefly narrates the decision of smog disturbing, image shake and light detection at night. The following two chapters discuss two application systems: video detection assistant warning system and speed measurement system.