Car License Plate Recognition (LPR) system has a broad application to the transportation management and the public security supervision. In this paper, a complete LPR system is addressed including the gate system and the algorithm on the license plate recognition. The gate system captures the images of the vehicles through a specified region using background subtraction and the LPR algorithm recognizes the characters of the captured vehicle. Especially in the character segmentation of the LPR algorithm, the concept of the similarity between a character string and license plate is introduced and based on this, a new algorithm on the character segmentation of license plate is proposed as a part of the LPR algorithm. And in the character recognition of the LPR algorithm, the 2-D discrete wavelet transformation and the radial basis function neural networks are used to recognize the characters segmented by the character segmentation algorithm. Experiment results show that the gate system and the LPR algorithm have good feasibility and effectiveness.
|