Design And Implementation Of The Car License Plate Recognition System Based On Fuzzy Theory

The techniques of License Plate Recognition (LPR), as one of the key techniques of intelligent transportation system, were extensively studied both on home and abroad. It has important theoretical significance and application value to Research the LPR system. The LPR system usually includes the processes of image pretreatment, plate location, character recognition.In this thesis, based on the characteristics of the License Plate, the complete research of LPR system were explored and investigated with fuzzy theory. The main results are as follows:1. The importance of LPR for ITS was pointed out. After reviewing the existing LPR techniques, the suitability of fuzzy theory for LPR system was analyzed, and a whole research scheme of LPR based on fuzzy theory was designed.2. The pretreatment technology of the license plate image are investigated, which includes gray processing, image enhancement, image binarization, etc. a novel image segmentation based on fuzzy C clustering was put forward. The proposed method has extensive application areas because of the segmentation with no sample monitoring.3. The characteristics of the license plate were investigated in detail. A new plate location method using color texture was built up, and the character split was carried out with the preset threshold.4. The characters were extracted with the method of coarse grid, and a fuzzy neural network classifier with four-layer structure was designed. The experiment was undertook in Matlab platform, the results indicated the propose method has good suitability for LPR.