摘要：随着经济的快速发展，机动车在生活中的作用日益显著，为了疏通车流，各地区都建立起了日趋复杂的交通网，而交通网的正常运行离不开信号灯的指示。另外，由于道路状况的复杂性和多样性，驾驶员可能会出现注意力分散等情况，从而导致交通事故的发生。信号灯识别系统作为一种有效的辅助驾驶系统，在智能交通生活中扮演着越来越重要的作用，是智能交通系统中不可或缺的一环。

本文围绕着交通信号灯提出了一种基于图像处理的检测与识别方法，并通过MATLAB平台进行实验研究。本文主要研究内容如下：首先，对交通灯进行信号检测，通过交通灯的发光特性，对信号进行颜色分割，由此确定信号灯的候选区域；其次，对交通灯进行识别，先用标记法对兴趣区域进行标记，再通过面积及圆形度等参数对其进行特征提取；最后，通过模板匹配法对处理后的图像进行匹配，最终得到交通灯的实时状态。

仿真结果表明，本文设计的交通灯识别系统基本能够满足设计要求，能有效地识别交通信号灯的状态，在很大程度上为城市交通安全提供了保障。

关键词：交通信号灯；颜色分割；特征提取；模板匹配

Abstract：With the rapid development of economy, the role of motor vehicles in life is becoming more and more obvious. In order to dredge traffic flow, every area has established a complex traffic network, and the normal operation of the traffic network cannot do without signal orderly grooming.On the other hand, due to the complexity and persity of the road conditions, drivers may be distracted and finally leading to traffic accidents.As an effective auxiliary driving system, the signal recognition system is playing an increasingly important role in the intelligent transportation and it’s an indispensable part of the intelligent transportation system.

In this paper, a method of detection and recognition based on image processing is put forward around the traffic lights,and the experimental research is carried out through the MATLAB platform. The main contents are as follows,the first is the detection of the traffic lights,by the light characteristics of the traffic lights,the color of the signal is segmented to get the candidate area of the traffic light.The second is the identification of the traffic lights, the region of interest is labeled by the mark method at first and then extract the features by the parameters of area and roundness.Finally, the template matching method is used to match the processed image and then obtains the real-time state of traffic lights in the end.

The experimental result shows that the traffic light identification system designed in this paper can basically meet the design requirements, and can effectively identify the status of traffic lights, what’s more, it provides a guarantee for urban traffic safety in a great extent.

Keywords:Traffic lights; color segmentation; feature extraction; template matching

目 录

第一章 绪论 1

1.1研究背景及意义 1

1.2国内外研究现状 1

1.3交通信号灯识别难点及研究安排 2

第二章 信号灯及图像处理基础理论 4

2.1交通信号灯的介绍 4

2.2彩色图像的介绍 5

2.3图像处理的基础知识 8

2.3.1图像预处理 8

2.3.2形态学处理