基于图像检测的交通信号机技术研究

【作者】 李哲；

【导师】 史忠科；

【作者基本信息】 西北工业大学 ， 交通信息工程及控制， 2005， 硕士

【摘要】 城市交通信号机是解决城市交通拥挤问题的重要工具。然而,目前大多数交通信号机采用了环形线圈等设备来检测交通参数信息,不管是从安装维护方面还是从技术方面来讲,都存在很大的缺陷。论文针对这个问题,对基于DSP图像检测的信号机进行研究和探讨。 论文首先介绍了图像交通信号机的整体结构,然后阐述了硬件电路各关键部分的设计思想,在此基础上着重分析了图像检测板电路的各部分硬件设计,并对高速电路板抗干扰设计采取的措施做了说明;在软件设计方面,对系统初始化、图像采集和处理程序等做了较详细的介绍,图像处理中采用了阈值分割和边缘分割的两种处理算法保证交叉口车辆排队长度检测的准确,并利用立体视觉测距原理,计算出了车辆排队长度的实际长度,为交叉口交通信号的实时控制提供保障:最后总结全文,并就进一步工作做了展望。更多还原

【Abstract】 The Urban traffic signal controller is one of the important parts of Intelligent Transportation System(ITS). Nowadays,most of the traffic singnal machine adopt the inductance coil and etc to obtain traffic information.However there are many serious problems in fixing, maintenance and technique. In order to deal with those problems above, the traffic signal controller based on the image processing by DSP have beenpresented in the paper.The whole structure of the traffic signal control machine based on imageprocessing is presented firstlly, then an idea of the circruitry to process image by DSP is put forward and the working means is detailed,furthermore every important part of circruitry is described. Not only did we realize the hardware circuits of the system, but also we developed it’s software. After the program of the system is described in detail ,the methods of threshold segmentation and edge segmentation are discussed emphatically, the length of the motor vehicle queue from the picture captured on traffic cross road is computed by the method of two detection methods. Finally, the main work of the paper is summarized, and suggestions are given for future study .更多还原