| In order to cope with shunting accidents because of human factors and improve the security of shunting operation, the paper constructs a shunting safety assistant system for classification yard based on computer vision and image processing, which is used to guide drivers to control locomotive correctly, and avoid shunting accidents, such as blue-light running, hustling railway switch and colliding earth retaining.The characteristics of railway shunting operation are analysed combined with perspective principles.The obstacle detection algorithm based on N-divided sub-sampling is proposed. In addition, the paper researchs the distance estimation between locomotive and targets by edge extraction and Hough transform.The precision of this algorithm can gain performance satisfactorily. A particular application of image stabilization is concerned in the paper. In order to verify its validity, it has been tested on a variety of color video-sequences taken from the point of view of trains moving along railways tracks.In the paper a general system for real-time detection and recognition of traffc signals is proposed. The key sensor is a camera installed on a moving vehicle. The software system consists of three main modules: detection, tracking, sample-based classification and recognition. Additional sensor information, such as vehicle data, GPS, and enhanced digital maps are used to enhance the performance and robustness of the system. Since the detection and recognition step is the most critical one, a detection and recognition algorithm based on shape and color features is proposed. Firstly, it makes use of the shape feature to locate the signal lamps,and then judges their status at HIS color space by the colors.The color system, with a high dynamic range camera and precise location information of the vehicle and the searched traffic signals, offers valuable and reliable help in directing the driver's attention to traffic signals and thus can reduce blue-light running accidents.
|
PDF Full Text Request