Vehicle License Plate Localization without Constraints

• Main Authors: Yen-Kun Lu, 呂炎坤
• Other Authors: Wen-Kuang Chou
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/59406457438839097408

碩士 === 靜宜大學 === 資訊管理學系研究所 === 95 === In decades, the Vehicle License Plate Recognition system has been discussed in many scientific papers, and also some relative products have been produced and applied. However, because of the different environments, the VPL recognition system hasn’t been used in mobile applications yet. The different environmental disturbances make vehicle plate localization system very difficult. The major issue lies in the complexity of environments. This research will develop a new approach to solve the complicated localization problem. To overcome the environmental constraints on shooting vehicle plates, we need a fast tracking method to deal with the massive images sufficiently and effectively. Therefore, we choose a simple partition method of the profile and modify the profile tracking method to the gray one. The proposed approach can find the correct location of vehicle plates in most conditions. The proposed profile tracking method on the gray images has the high recognition rate even with interference. Most importantly, the system is also efficient and effective. Our experiments show that the system takes 300 ms in average to localize a plate with 94.75% correct rate, even various images in different environments are tested.