Embedded System Implementation for License Plate Recognition

• Main Authors: Hsin-Tsung Feng, 馮信璁
• Other Authors: 陳敦裕
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/71762219451264254073

碩士 === 元智大學 === 電機工程學系 === 98 === In this paper, we present a license plate recognition system in the embedded system platform. Our system employs the Ada-Boosting technique to train the model in order to detect the license plate. After the license plate is detected, this system will segment the license plate into some characters and will extract the features. Subsequently, we use the SVM classifier to classify the feature classes so as to recognize the accurate character’s meaning. Finally, these meaningful characters are checked by the mechanism of syntax analysis in order to differentiate the false results. Our embedded system DaVinci 6446 platform is composed of the ARM and DSP units, which is manufactured by the Texas Instruments. DaVinci 6446 platform is a two-core processor. ARM is responsible for the system control and DSP processes the complex mathematical operation. By the cooperation of the two cores, the performance can achieve real-time processing. Our system obtains an average processing time of 41ms per frames, about 25 fps.