Pyramidal approach to license plate segmentation

摘要: Car identification is a goal in traffic control, transport planning, travel time measurement, managing parking lot and so on. Most car algorithms contain standalone plate segmentation process followed by contents reading. A pyramidal algorithm for license segmentation, looking textured regions, has been developed on PC based system running Unix. It can be used directly applications not requiring real time. When input images are relatively small, real-time performance fact accomplished the algorithm. using large images, porting to special digital signal processors easily lead preserving performance. Experimental results, stationary moving cars outdoor scenes, showed high accuracy scores detecting plate. The also deals with cases where many character strings present image, only one corresponding This done means of constrained texture regions classification.