毕业论文翻译汽车车牌识别系统

车牌识别系统是计算机视频图像识别技术在车辆牌照识别中的一种应用。车牌识别在高速公路车辆管理中得到广泛应用，电子收费（ETC）系统中，也是结合DSRC技术识别车辆身份的主要手段。

车牌号码识别系统是基于车牌识别技术而衍生的一个完整系统。首先将运动中的汽车牌照从复杂背景中提取出来，通过车牌提取，图像预处理，特征提取，车牌字符识别等技术，识别车辆牌号，输出车牌号码。技术水平为字母和数字的识别率可达到96%，汉字的识别率可达到95%。

在停车场管理中，车牌识别技术也是识别车辆身份的主要手段。在深圳市公安局建设的《停车库（场）车辆图像和号牌信息采集与传输系统技术要求》中，车牌识别技术成为车辆身份识别的主要手段。

车牌识别技术结合电子不停车收费系统（ETC）识别车辆，过往车辆通过道口时无须停车，即能够实现车辆身份自动识别、自动收费。在车场管理中，为提高出入口车辆通行效率，车牌识别针对无需收停车费的车辆（如月卡车、内部免费通行车辆），建设无人值守的快速通道，免取卡，不停车的出入体验，正改变出入停车场的管理模式。

一、系统简介什么是车牌识别智能称重系统呢?相信许多没有在运输，大众货物交易等环境工作的人们，都不是很明白。智能称重管理系统需要在测量车辆载重的同时记录车牌号码，并建立对应关系。采用人工录入车牌方式，需要较长的等待时间，外部工作环境比较恶略，占用较多的人员成本，且容易产生失误甚至作弊等问题。应用高清车牌识别相机产品，在车辆称重的同时自动识别车牌/及其车牌号码，同时记录车辆图像、检测数据等信息。大大节省了检测时间，提高了工作效率且有效避免人为因素影响。车辆信息还可自动上传到管理中心实现联网操作。结合称重系统，高清车牌识别相机产品可广泛应用于公路治超检测站、高速动态称重、企业超限治理、称重付费等领域车牌自动识别产品的应用实现了称重管理系统的智能化。通过车牌号码的自动识别、自动登录、自动比对，系统可以实现自动抬杆、自动验证用户身份、超载车辆自动报警等诸多智能化功能。智能化有效提高了称重管理系统的管理质量，将获得良好的经济效益和社会效益。二、系统特点MAXFORT系列牌照识别仪区别于传统的工控机体系，采用了DSP为核心的嵌入式系统体系，一路视频一个工作单元的模块化结构以及网络化、TCP/IP协议的数据传输方式，无需工控机即可实现自动检测车辆、获取车辆信息，将数据通过网络发往系统中心。采用MAXFORT系列牌照识别仪来采集车辆信息的称重管理系统具有以下特点：采用车牌自动识别，减轻了人工手动录入车牌信息，缩短了车辆在称上等待时间，提高了服务质量。优秀的识别算法、智能调光技术，全天候的高识别率保障了系统的性能。嵌入式识别方式，结构、功能模块化，新系统实现简单，方便原有系统改造。识别设备独立工作，只将识别结果发给管理计算机，工作稳定、可靠性高。MAXFORT系列牌照识别仪具备优秀的视频检测车辆的功能，可以与埋地线圈方式配合使用，线圈故障时可自动启用视频检测功能。

A license plate recognition system can be divided into three steps: license plate detection, character segmentation and character recognition. License plate detection will be inclined, light, pollution and so on, algorithm complexity. In recent years, put forward a lot of license plate detection algorithm, but most of them are inadequate, not in a complex environment, achieve a good license plate recognition. Edge detection algorithm with filming the impact of changes, when there is no clear license plate edge can be detected in the plate. Combination of color and texture feature detection accuracy is better, but by the influence of illumination is very obvious, and the algorithm complexity is high, no effect. The morphology of license plate detection algorithm, it is easy to be affected by class of license plate character effects, detection speed is slow. These detection algorithm has shortcomings because of lack of adaptation algorithm. This paper presents the Boosting algorithm in the AdaBoost algorithm, this is a kind of structure accurately classifier adaptive learning algorithm, it combines many features to construct accurate classifier, finally complete license plate paper studies the key technologies in license plate recognition, license plate location, introduces in detail the character segmentation, feature extraction and recognition method of license plate location, and with the AdaBoost algorithm based on license plate recognition is discussed, the results show that the AdaBoost algorithm can evidently improve the stability and precision of license plate detection of.

A license plate recognition system can be divided into three steps: license plate detection, character segmentation and character recognition. License plate detection will be inclined, light, pollution and so on, algorithm complexity. In recent years, put forward a lot of license plate detection algorithm, but most of them are inadequate, not in a complex environment, achieve a good license plate recognition. Edge detection algorithm with filming the impact of changes, when there is no clear license plate edge can be detected in the plate. Combination of color and texture feature detection accuracy is better, but by the influence of illumination is very obvious, and the algorithm complexity is high, no effect. The morphology of license plate detection algorithm, it is easy to be affected by class of license plate character effects, detection speed is slow. These detection algorithm has shortcomings because of lack of adaptation algorithm. This paper presents the Boosting algorithm in the AdaBoost algorithm, this is a kind of structure accurately classifier adaptive learning algorithm, it combines many features to construct accurate classifier, finally complete license plate paper studies the key technologies in license plate recognition, license plate location, introduces in detail the character segmentation, feature extraction and recognition method of license plate location, and with the AdaBoost algorithm based on license plate recognition is discussed, the results show that the AdaBoost algorithm can evidently improve the stability and precision of license plate detection of.（望采纳）嘿嘿～～

A license plate recognition system can be divided into three steps: license plate detection, character segmentation and character recognition. Measure the main will be affected by license plate tilt, illumination, pollution and so on, the effects of the algorithm is more complex. In recent years many license plate detection algorithm is put forward, but most of them have enough to under the complex environment well realize real-time identification of the license plate. Edge detection algorithm with shooting influence change, when there is no clear license plate edge to detect the license plate can not succeed. Color and texture features comprehensive detection algorithm is better accuracy, but by light effect obviously, algorithm complexity and high, the effect is not. For morphological license plate detection algorithm, it is easy to be kind of license plate character, the effects of the detection speed is slow. Above these detection algorithm has shortcomings because of lack of ability to adapt to the algorithm. This paper puts forward the delegate AdaBoost algorithm Boosting algorithm, this is a kind of tectonic accurately classifying adaptive learning algorithm, it through the combining multiple features to construct the precise classifier, finally complete license plate detection. This paper studies the license plate identification key technology, detailed introduces the license plate localization, character segmentation, feature extraction and character recognition method, and the license plate location and AdaBoost algorithm based on the license plate identification were studied, and the results show that AdaBoost algorithm obviously improving the license plate detection stability and precision of sex.