Research Of License Plate Recognition System Based On Distributed Character Of Projection Image And Neural Network

The license plate recognition is an important link in intelligence transportation system. It holds the very important status in the transportation surveillance and the control, the automatic license plate recognition system can used to the field of transportation information statistics and collecting, correspondence between road and car, parking lot management, charging without stopping car, vehicles automatic travel, can used to register the passing vehicles in real time, counter current capacity and garrison the defensive goal in the transportation street intersection, the highway, the military fort and the institution gate guard, can recognize and hold up the vehicles which cause troubles, the vehicles that are robbed and the criminous vehicles;it also can used to register, counter and inquiry the vehicles which pass in and out of the automobile parking lot, the article, combine with reality, has carried on research to license plate locatication, character segmentation and character recognition, the main result of research is as follows:Firstly, it analyzes several traditional plate locating methods, presents a license plate locating algorithm based on the distributed character of projection image and mathematical morphology, namely, using the horizontal gradient algorithm to get the horizontal projection image according as the characteristic of rich vehicle plate regional veins, then using the frequency of gray scale which is in the image to find the horizontal location of vehicle plate, sith, using the opening and closing operations of the mathematical morphology to do vertical location of the vehicle plate, so it adjusts parameters of the texture analysis algorithm.Second, it analyzes several algorithms of geometric correction. Furthermore, it presents a method of improving Hough transform, this method can correct the gradient of license plate images accurately. It designs an algorithm of combing with unitaring and projection character which based on the characteric of license plate to achieve the separation.For the typical difficulties in license plate image character recognition that the character blur, character destruction, character image stained because the effects of vidicon's performance, license plate's neatness, illumination etc. So arithmetic must have good adaptability. The system use BP neural pattern recognition. The BP neural pattern recognition is considered as the recognition method and vehicles' license plates is taken as recognition object, which is to improve the recognition rate of the character recognition of vehicles' license plates under noising circumstance. In the end I get a good license plate recognition result.On the basis of arithmetic research, this article have realized the relative algorithm and the interface of the license recognition software with the programminglanguage of Visual C++, and it construct a practical license recognition, control the administrative system, this system can read and save the license image and the vehicle information, recognize the license character, transmit the vehicle information and image, do database operating of vehicle information.The studies of the article indicated that, the license plate recognition method which uses the distribution characteristic of projection image and the nerve network can effectively enhance the anti-jamming and the recognition rate of the system. The experiments show that this system has high rate of accuracy, quick speed of recognition, good robustness, and better practical value.