Research And Application Of Vehicle License Plate Recognition System

Vehicle license plate recognition system is an important technique in intelligent traffic system. It plays a key role for vehicle automatic management in intelligent traffic system. Vehicle license plate recognition system will be more widely used based on driving cars, and became an important part of intelligent transportation systems because of its adaptability, high degree of automation.Dynamic vehicle license plate recognition system were studied in this thesis,and complete the following work:1.License plate image preprocessing method was presented, vehicle detection and localization are studied; the vehicle license plate image after preprocessing were introduced. Ashy, filtering, image enhancement, image binarization and a series of operations and processing, can improve the image quality, for vehicle license plate location and character segmentation and recognition of image processing speed and accuracy, laying the foundation to improve the performance.2.The thesis researched vehicles license of positioning method, and positioning algorithm, and character segmentation method and character recognition of method in detail. The thesis proposed has paper used to border segmentation for based, based on vertical projection of character segmentation method, the method simple effective of license plate positioning, and character segmentation and recognition of method, needs between character and character of clearance at get character in vertical direction up of projection of local minimum value, according to minimum value of location segmentation out character; In terms of character recognition using neural networks based template matching method for character recognition.3.This paper presents a method based on improved neural networks and an improved template matching method for character recognition. To design the Shenyang renovation project on the management of a shopping centre car park, vehicle license plate recognition system for management were used in the underground car park. The algorithm is realized in this system and through experiment simulation validation the effectiveness, its advantage is to overcome the template matching numbers, letters and characters looking for difficult shaped word images, respectively, and using neural networks to identify the characters, test results show that the character recognition rate is increased and achieved good results.4.hicle license plate character recognition based on template matching and neural network algorithms were proposed. It nicely solves the problem of inaccurate font adhesions, segmentation, and more than ever with high recognition rate.