车辆图像智能抓拍系统的设计与实现

发布时间：2018-06-26 05:23

本文选题：智能交通 + 车辆通过检测　；参考：《电子科技大学》2014年硕士论文

【摘要】：图像抓拍技术(Video Enforcement System,VES)是智能交通系统的重要组成部分。本文通过车辆检测技术、图像抓拍技术以及车牌自动识别技术,自动捕获路口红灯时所经过的车辆车牌信息,从而实现道路交通的智能管理。本文对车辆图像智能抓拍系统的研究主要集中在如下几个方面。(1)车辆通过检测子系统通过红灯检测器、环形感应线圈和Athenex车辆自动检测器,实现对路口红灯信息的自动监控,当信号灯状态为绿灯状态时,关闭车辆通过检测子系统,;当信号灯为红灯状态时,当检测到有车辆通过时,启动图像抓拍模块,进行通过车辆的抓拍。通过对红灯信号的检测,实现车辆检测模块的自动开关,达到降低误拍、节省能源的目标。(2)图像抓拍功能子系统图像抓拍功能子系统在接收到车辆检测功能子系统所发送的拍摄信号之后,进行拍照,并且将所拍得的图片发送到服务器端进行处理。采用工业摄像机,在车辆通过检测模块的控制下,进行闯红灯车辆的抓拍,并且将抓拍的图像,通过Socket技术传输到服务器端进行车牌号码的自动识别。(3)车牌号码自动识别子系统车牌号码自动识别功能是在接收到图片抓拍功能子系统上传的闯红灯车辆抓拍图片之后,采用车牌号码自动识别技术识别抓拍图片中的车牌号码,并且将车牌号码、时间、地点等信息传送给交通管理系统。在本文的研究中,首先通过对抓拍图像的分析,进行车牌号码的定位,并且利用Tury LRP车牌号码自动识别技术,首先对车牌号码的自动识别。(4)信息发布子系统车辆图像智能抓拍系统出了对闯红灯的车辆进行抓拍,并且进行抓拍图像的车牌号码自动识别以外,更为重要的是为城市交通管理系统提供闯红灯车辆的车牌号码等信息。而为了满足系统可扩充性的需求,以及为了满足不同操作系统平台的要求,本文所研究的车辆图像智能抓拍系统采用SOA技术来实现数据的共享。在测试中,系统对闯红灯车辆的抓拍率达到98%以上,车牌号码的自动识别率达到95%以上,人工识别率达到100%。表明本文所设计的车辆图像智能抓拍系统能够满足路口闯红灯车辆的智能抓拍要求,符合系统设计时的要求。
[Abstract]:Video Enforcement System (VES) is an important part of the intelligent transportation system. In this paper, the vehicle license plate information is automatically captured by vehicle detection technology, image capture technique and license plate automatic recognition technology, so as to realize the intelligent management of road traffic. The intelligent grasp of vehicle image is carried out in this paper. The research of the beat system is mainly concentrated in the following aspects. (1) the vehicle through the detection subsystem through the red light detector, the ring induction coil and the Athenex vehicle automatic detector to realize the automatic monitoring of the red light information of the intersection. When the state of the signal light is green, the car is closed through the detection subsystem, and when the signal lamp is a red light state. When the vehicle is detected, the image capture module is started and the capture of the vehicle is carried out. The automatic switch of the vehicle detection module is realized through the detection of the red light signal to achieve the goal of reducing the error and saving energy. (2) the image capture function sub-system of the image capture function subsystem is sent to the vehicle detection function subsystem. After the shooting signal, take pictures, and send the pictures to the server to handle. Use industrial video camera, under the control of the vehicle through the detection module, carry on the capture of the red light vehicle, and transfer the captured image to the server side by Socket technology to automatically identify the number of the license plate. (3) license plate. The automatic recognition function of the number automatic identification subsystem is to identify the license plate number in the captured picture by using the license plate number automatic identification technique after receiving the picture of the red light vehicle captured by the picture capture function subsystem, and transmit the license number, time and ground point to the traffic management system. In the study, first of all, the location of the license plate number is carried out through the analysis of the captured image, and the auto recognition technology of Tury LRP license plate number is used to automatically identify the number of the license plate first. (4) the vehicle image intelligent capture system of the information publishing subsystem has captured the car of the red light, and the license plate number of the captured image is carried out. In addition to dynamic recognition, it is more important for the urban traffic management system to provide information about the license plate number of red light vehicles. In order to meet the requirements of the system extensibility and to meet the requirements of different operating system platforms, the vehicle image intelligent capture system studied in this paper uses SOA technology to realize data sharing. In the system, the rate of capturing the red light vehicles is up to 98%, the automatic recognition rate of the license plate number is above 95% and the artificial recognition rate reaches 100%., which shows that the intelligent capture system designed in this paper can meet the requirements of the intelligent capture of the red light vehicles at the intersection, and the system meets the requirements of the system timing.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U495;TP391.41

【相似文献】