【作者】 李晓飞；

【导师】 龚建荣；

【作者基本信息】 南京邮电大学 ， 信号与信息处理， 2013， 硕士

【摘要】 在智能交通系统中，视频分析技术可自动判断出交通视频中越线、逆行、闯红灯等违规行为。然而，道路监控网络的规模庞大，交通摄像头数量众多，监控器一旦出现断电黑屏、遮挡、雪花噪声或者由于器件老化、人为破坏引起的抖动、色偏、条纹等情况，将无法采集有效的视频信号。那么依赖于视频分析技术的智能交通系统也将无法正常工作。因此，对视频图像质量智能检测的研究有着重大意义。针对图像质量智能检测问题，本文设计了以达芬奇架构DSP芯片为核心的图像质量智能检测系统。首先完成了图像质量智能检测系统的总体框架的设计，确定了系统的设计目标和系统芯片的选型。然后实现了视频输入输出模块、DDR2、NANDFLASH存储模块和以太网物理层电路模块等系统电路的搭建，并且对高速电路模块进行了完整性分析，同时给出了高速电路问题的解决方案和系统仿真。最后研究了图像质量智能检测算法，并分析了算法在达芬奇架构上的移植和优化。本文在设计该系统时，充分考虑了高速电路设计问题，仿真结果表明其硬件电路性能稳定。另外该系统核心处理器通信接口丰富，并且采用LINUX操作系统，便于实现功能扩展和软件升级，可以满足智能交通中对低质图像进行智能检测并及时报修需求。更多还原

【Abstract】 In intelligent transportation system, the application of video technology in traffic behaviortest include: Cross the line, Retrograde, Running red lights and so on. However, intelligenttransportation system must be a huge Internet network. In the huge intelligent transportationsystem, the number of video monitor is a very astonishing number. If one of those videomonitors can’t provide normal video signal as a result of power outage, Black screen, shelter, snownoise or shake, color cast, streak caused by device aging.Then, the intelligent transportationsystem relies on video analysis techniques will not work.So the research of the intelligentdetection of image quality is of great significance.According to the intelligent detection of image quality, this paper completes the intelligentdetection of image quality system whose core processer is the TI Davinci framework DSP chip,Firstly,this paper completes the overall functional designing, the specific design goals’determination and selection of the system hardware chip.Secondly,this paper realizes the circuitdesign include: the video input and output modules,DDR2and NANDFLASH storagemodules,Ethernet physical layer modules and so on.It also simulates the results.At last,this papersets up the system software environment. Additionally, this paper designs and implements theimage quality intelligent detection algorithm and analyzes the transplantation and optimization ofit.In the design process of the system, we give full consideration to the high-speed problem ofit.The simulation results shows that it’s performance is stable.That the core process chip of thesystem not only has rich communication interfaces but also has LINUX operating system, so it iseasy to implement function expansion and software upgrade for the system.Because of that thesystem can be applied to the intelligent transportation system in device self-test and requestingrepair.更多还原