基于FPGA的交通图像检测系统研究与设计

发布时间：2018-03-01 23:07

本文关键词： 智能交通 FPGA 软件仿真图像处理　出处：《昆明理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：当今社会经济发展迅速,各种交通工具的数量大幅度增加,这也导致了城市交通密度的持续增加,各种交通事故层出不穷,城市交通水平越来越受到社会各界人士的关注,传统的依靠人力来协调和指挥交通已经无法满足需求。随着数字图像处理技术的发展,基于图像处理技术的智能交通管理系统也成为时下社会的研究热点之一。智能交通管理系统可以有效的帮助城市缓解道路拥堵、降低事故发生概率,有效地提高交通效率,近年来已经倍受重视。本文基于FPGA交通图像检测系统,首先分析了 PC构架的交通视频检测系统的弊端,其次,本文从稳定性、可靠性、开发经济性以及效率等方面的考虑,结合软件仿真分析,提出以FPGA作为图形处理器的设计方案。利用FPGA实现实时图像处理,在很大程度上增加了车辆检测的方便性和灵活性。本论文主要从以下两点进行研究:1、软件仿真。针对现有道路交通图像的实际情况以及特点,为了提高后续交通处理的准确度和方便硬件设计,利用OpenCV工具对交通图像的预处理、降噪、运动车辆检测和图像边缘检测进行了软件的仿真以及验证,使用更合理、简便的图像处理方式。2、基于FPGA的系统实现方面。根据论文第一部分的软件仿真分析结果,以及FPGA自身特点,设计了由交通图像数据采集模块,交通图像数据存储模块和交通图像处理模块组成的交通图像检测系统。该系统较好地解决了图像采集、图像分析等问题,为以后智能交通系统的功能拓展奠定了基础。本文所有交通图像处理算法均在OpenCV中进行仿真验证,并在FPGA中进行了实现。结合交通系统要求的可靠性与实时性,本文涉及到的FPGA各个功能模块代码用Verilog语言写成,最后经过系统的板级调试表明,能较好地完成交通图像处理与检测等工作,而且本文的研究内容对其他场合的图像处理及检测技术也有一定的参考价值。
[Abstract]:Nowadays, with the rapid development of social economy, the number of various kinds of means of transportation has increased substantially, which has also led to the continuous increase of urban traffic density, the emergence of various traffic accidents in endlessly, and the increasing concern of people from all walks of life about the level of urban traffic. The traditional reliance on manpower to coordinate and direct traffic can no longer meet the demand. With the development of digital image processing technology, The intelligent traffic management system based on image processing technology has become one of the hotspots in the society. Intelligent traffic management system can effectively help cities alleviate road congestion, reduce the probability of accidents, and effectively improve traffic efficiency. In recent years, more and more attention has been paid to the traffic image detection system based on FPGA. Firstly, the disadvantages of the traffic video detection system based on PC are analyzed. Secondly, the stability, reliability, development economy and efficiency are considered in this paper. Combined with the software simulation analysis, the design scheme of using FPGA as graphics processor is put forward. Real time image processing is realized by using FPGA. To a great extent, it increases the convenience and flexibility of vehicle detection. This paper mainly studies the following two aspects: 1, software simulation. According to the actual situation and characteristics of the existing road traffic images, In order to improve the accuracy and convenient hardware design of the subsequent traffic processing, the traffic image preprocessing, noise reduction, moving vehicle detection and image edge detection are simulated and verified by using the OpenCV tool, and the use is more reasonable. The simple image processing method. 2. Based on the system implementation of FPGA. According to the results of software simulation and analysis of the first part of the paper, and the characteristics of FPGA itself, the traffic image data acquisition module is designed. The traffic image detection system is composed of traffic image data storage module and traffic image processing module. All the traffic image processing algorithms are simulated in OpenCV and implemented in FPGA. The code of each function module of FPGA involved in this paper is written in Verilog language. Finally, the system board debugging shows that traffic image processing and detection can be completed well. And the research content of this paper also has certain reference value to the image processing and detection technology of other occasions.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【参考文献】