Research And Implementation Of License Plate Recognition Technology

License Plate Recognition (LPR) is an important study field which the computer vision and pattern recognition are used in the Intelligent Transportation Systems (ITS), and the LPR is the key technology to realize the ITS. The typical LPR system consists of four modules in general; those are the pretreatment of image, license plate location, character segmentation and character recognition. LPR is studied from these four steps in the thesis, and a series of effective algorithms are used in LPR based on experimental results. Each of aforesaid is implemented with computer program, and the result is acceptable.The pretreatment is needed before locating license plate in order to eliminate the noise, improve the quality of image and make the area of License Plate intensified. I- mage pretreatment includes image gray, image binaryzation, median filtering and ed- ge detection.In license plate location, it analyzes and compares many license plate location a- lgorithms, such as edge detection method, artificial neural network method, gray me- thod and method base on color, etc. It brings forward a method of combining the ed- ge detection and line scan to locate the license plate. It satisfies the demand in speed and accuracy.In character segmentation, it introduces a lot of algorithms about license plate image binary, geometrical revising and standardization. And then a character segmentation method is used to segment character, which is base on projection pre-known knowledge of the size and the space between characters of vehicle license plates, and has a good effect.In character recognition, it analyzes and compares two ordinary methods about license plate character recognition. And then it puts forward thoughtway of combination of character recognition algorithm. It adopts the different techniques to recognise Chinese characters, letters and figure characters. It recognizes the Chinese characters by the method of supporting vector machine; And to the letters and figure characters, it proposes a novel algorithm in character recognition based on character configuration, after analyzing the structural properties of letters and figure characters. It can get higher distinguishability through this method.The algorithm has strong anti-jamming ability to lean, distort and blur. Especial- ly, the method is robust to images with poor illumination or complex background.