Research On Automobile License Plate Recognition System

The automatic license plate recognition is an important technique to obtain traffic information, it mixes computer-vision,image processing techniques and pattern recognition techniques, it is an important technique in intelligent traffic system. It has a significant meaningfulness for ITS(Intelligence Transportation System), ETC(Electronic Toll Collection), regional vehicle administration and depot vehicle administration.LPR system contains two parts: license plate locating and license plate recognition. In researching the algorithm of license plate locating, first, preprocessing the image, including graying the RGB image,enhancing the gray image and transforming the gray image to binary image, then analyzing the characters of Chinese car, and detecting the license plate location use the method which based on license plate projection character and property of license plate. The license plate recognition has two parts: license plate characters partition and characters recognition. We emendate the license plate at first in the characters segment part, then utilize the qualify of characters such as the width of characters and the space between characters to segment the characters. The method is simple and applied, and the result of partition is good. In the part of the characters recognition, we make the character to the same scale at first, then get the stamp of the characters through the method of rough lattice and the method of the projecting density. At last we emphatically discusses the theory principle of the neural network and the application method in character recognition with it, and using the neural network to recognize the license plate characters.As the test result, it proves that this recognition system can relatively accurate to locate license plate, segment license plate and recognize the license plate characters, and also it has strong anti-jamming ability to lean and distortion.