【作者】 张航；

【导师】 罗大庸；

【作者基本信息】 中南大学 ， 交通信息工程及控制， 2006， 博士

【摘要】 随着城市化的进展，汽车的普及，机动车数量、出行人数的大量增加，路网通过能力难以满足交通量快速增长的需要，交通拥挤加剧，交通事故频发，公路交通的安全以及运输效率问题变得日益突出，智能交通系统(Intelligent Transportation System，ITS)被认为是解决这些问题的基本手段，ITS是一个集通讯、检测、控制和计算机技术于一体的综合系统，对保障交通系统的运行安全及运输效率，促进国民经济的发展有重要的意义和经济价值。目前，智能交通系统在我国发展迅速，且已正式列入“十一五”发展规划，驾驶安全问题、城市交通堵塞问题、运输效率问题都有望通过对车辆信息化和智能化的改造获得改善，基于图像检测与处理技术的交通标志识别系统作为智能交通系统的一个重要的子系统，已逐渐成为目前智能交通系统国内外研究的热点。目前，已有的各种交通标志识别算法，各具特色，在某些特定场合发挥一定的功效，不过也存在一些缺陷，因此，采用新的技术和理论，克服这些不足，提高算法的实时性和精度，对算法的实际应用具有重要的意义，同时，它作为模式识别领域的一个典型问题，也必将推动模式识别的理论和技术的进一步发展。 交通标志识别是通过安装在交通工具上的摄像机摄取户外自然场景中交通标志图像，输入计算机进行处理完成的，由于户外环境存在多种复杂因素的影响，它比一般的非自然场景下的目标识别更具挑战性，交通标志识别的主要难点有两点：1)是比较满意图像的实时获取；2)是上百种标志库的匹配和理解。本学位论文结合中国的实际，针对这两个难点，在交通标志的侦测算法、交通标志图像的盲复原算法以及交通标志的分类算法等三方面作了一些有益的工作，学位论文的主要工作和成果可以概括为以下五个方面： 1、在综述交通标志侦测算法、交通标志预处理算法以及交通标志分类算法的国内外研究现状的基础上，结合模式识别的基础理论和技术，提出了一种能并行处理、多层结构的交通标志识别框架，该框架具有结构简单、易于实现、实时性强的特点。 2、针对目前交通标志侦测算法的不足，深入研究了模板匹配技术在交通标志侦测中的应用，综合改进的粒子群优化搜索算法、结合更多还原

【Abstract】 With the development of city planning and the popular of the cars, the number of running cars and traveling men is greatly increasing. Because the existed road cannot meet the demand of traffic flux, the problems such as traffic jams, accidents, road safeties and efficiencies in transportation are becoming more and more serious. Intelligent Transportation System (ITS) is an important means to solve these problems, which is an integrated system combining the technology of communication, measurement, control and computer. It is of significance to guarantee road safety, improve transportation ability and promote the development of economics. At the present time, ITS has been developed greatly in our country and been listed in the eleventh five-year development plan. As an important system in ITS, the detection and recognition of traffic signs from natural scenes is becoming one of the hot spots in the international research field of the Intelligent Transportation System, which based on the technique of image processing. Although many algorithms for recognition of traffic signs are developed and they play the important roles in some practice instances, they also have some essential drawbacks and deficiencies. So it is important to study the refinement algorithms with the more real-time and better precision using the new developed techniques and theories for their applications. At the same time, as a type problem in pattern recognition, the study of the recognition of traffic signs can make progress in the field of pattern recognition both at theories and techniques.Traffic signs from natural scenes are received from the video on board and to be put to computers. Then through the image processing using the developed algorithms in computer, traffic signs can be recognized. But the study of recognition of traffic signs from natural scenes is more difficult and with more challenge than the study in non-natural scenes. There are two main difficulties in the recognition of traffic signs from natural scenes, including 1) how to get the satisfying traffic signs images更多还原