机构地区: 华南师范大学计算机学院计算机工程系
出 处: 《西安航空技术高等专科学校学报》 2007年第1期43-45,共3页
摘 要: 车牌识别技术(LPR)是智能交通(ITS)领域的重要研究课题之一。车牌的自动识别分为车牌检测、字符分割、字符识别三个主要部分。车牌检测是LPR中的关键技术。车牌图像由于受到噪声、车牌倾斜、光照不均、运动模糊等的影响,检测存在着很大的困难。因此,在总结前人检测算法的优点和缺点的基础上,提出了基于adaboost算法的检测算法,其优点是检测率高,检测时间少,适应性特别强。 License Plate Recognition (LPR) proved an crucial researching issue in Intelligent Traffic System( ITS). The LPR's automatic recognition consists of three sections: license plat detection, character segmentation and character recognition. The license plate detection turned out to be the key technology in the LPR, plate image's detection exist perplexities due to the noise , plate inclination, illumination variance, motion illegibility. Hence, the essay summarized merits and demerits of ascendants based on computation, proposing a new detecting approach based on adaboost algorithm , the advantage of which revealed high detection rate, less time consuming and chronic adaptability.
领 域: [自动化与计算机技术] [自动化与计算机技术]