（字数：30776  页数：66）摘 要：本文以视频检测技术、图像处理技术和数据传输技术为依托，以“交通道口信号检测与数据处理系统研究”为课题，对交通道口车流量信号的检测和数据传输技术进行了研究。论文首先介绍了交通道口信号检测与分析的重要性，以及国内国外的研究、应用现状，给出了这项研究的重要意义和前景。其次对多种车辆信号检测方法优缺点做了比较，选取了现阶段最优的信号检测方法进行研究。再次，对图像滤波除噪、图像锐化、图像增强、图像分割、特征提取与目标识别等图像处理过程中遇到的问题进行了分析比较，并在比较分析的基础上，确定了系统设计和实现中所要采用的基本算法。最后，对检测到的信号进行处理和传输方法进行理论研究。
本论文的主要特点是:
1、对交通道口上车辆信号（车型、车速、流量、特征）的各种检测方法进行分析、比较、优化，包括磁感应检测器、磁阻传感器、微波检测器、红外线检测器、超声波检测器、视频图像处理检测等方法，并且选出最优的检测方法视频检测方法。
2、对视频信号检测中的各种图像处理技术，包括图像预处理技术、图像分割技术、特征提取技术、图像识别技术进行分析。
3、在目标识别过程中提出了一种通过监督学习来训练判别函数的方法，通过对图像样本作某种监测，使它与输入图像进行比较，达到对输入图像进行分类的目的，并给出了相应的实现算法。
4、对信号检测过程中数据的传输和特征提取及图像识别处理的研究，制定基于RS-485总线的通信协议。
开展交通道口信号检测和数据处理系的研究，实时了解城市交通拥挤道路交通状况，并且及时将了解到的交通道口信息进行分析，便于广泛开展城市交通的理论研究，进行各种工程措施、管理措施实施情况的前后对比，判断改善交通措施的效果等。交通道口视频检测技术的研究，对于提高我国交通信号检测技术水平，促进城市道路建设，实现城市交通管理智能化都具有十分重要的现实意义。

关键词: 信号检测,交通流量,车辆识别,图像处理,特征提取,数据传输,通信协议


Abstract
Basing on the task of Traffic Road Junction Signal examination and Data processing System based on the technology of Video examination, image processing and the Data transmission. It researches into the examination of Traffic road junction flow and the technology of Data processing. It emphasized in the following aspects. First, the importance of traffic Road Junction Signal examination and analysis, as well as the current research and application of this technology in our country and abroad, the significance and the prospect of this research are provided. Second, comparison with different kinds of vehicles signal examination method that has advantages and disadvantages are given in this thesis, what’s more , selection of the superior signal examination method of present stage to conduct the research is also introduced. Third， comparison and analysis of the problems that exist in image processing，such as image filtering and noise reduction，image sharpening，image enhancement，image segmentation，feature pick—up and object recognition are shown. Thus，it establishes the basic method of calculation that will be adopted in the design and realization of the system.
The major characteristics of this thesis are as follows:
1. Basing on the analysis, comparison and optimization of the different methods of Vehicles signal examination (Vehicle type, vehicle speed, current capacity, character), including the examination method of magnetic induction detector, magnetic resistance sensor, microwave detector, infrared detector, ultrasonic wave detector. Meanwhile, it presents the best method of Video examination.
2. Emphasizing on the analysis of video signal examination of different kinds of image processing technology, including the technology of image pretreatment, the image division , the characteristic extraction , the pattern recognition .
3. Target identification process put forward a method of supervised learning to train the decision function and monitor the image sample, comparing with the input image and achieving the goal that classifies the input image and provides the corresponding method of calculation to realize it.
4. The research on the transmission of signal examination data , the character pick up and pattern recognition processing base on the RS-485 main line correspondence agreement
At Present，researching into the traffic road junction signal examination, comprehending the real-time crowded road traffic condition of the city that has comparison with each kind of project and management measure , judging the effect of improvement transportation measure and so on, which can analyze the information in time and widely develop the theoretical research on the city transportation . It is of great of significance to improve the level of traffic signal examination technology , promote the urban road construction in our country and realize the intellectualization of traffic management .

Key words: Signal examination; Traffic flow; Vehicles recognition; Image processing; character pick up; Data transmission; Communication protocol.

目录

第1章 绪 论 1
1.1 概论 1
1.2 国内外现状及现阶段发展方向 2
1.3交通道口车辆检测的方法 3
1.4 论文的主要任务 9
1.5 论文的组织结构 10
第2章 目标检测和目标识别的基本原理 12
2.1 目标图像输入 12
2.2 图像预处理 13
2.3 图像分割与特征提取 14
2.4 目标检测与识别 14
第3章 检测到的图像预处理 15
3.1 数字图像 15
3.2 图像噪声的消除 18
3.3 图像锐化 21
3.4 对比度的增强 22
3.5 小结 26
第4章 图像的分割与特征提取 27
4.1 图像分割 27
4.2 特征提取 30
4.3 小结 33
第5章 交通道口目标的识别 34
5.1 特征选择 34
5.2 判别函数 35
5.3 监督学习 37
5.4 系数调整(误差修正) 38
5.5 小结 39
第6章 交通道口信号检测与数据处理系统的设计 40
6.1 系统概述 40
6.2 系统设计原则 40
6.3 系统总体规划 42
6.4 系统软件功能模块设计 44
6.5 小结 48
第7章 交通道口检测到信号的传输 50
7.1 性能要求 50
7.2 数据通信协议的设计方案 51
7.3 据通信协议实现流程 54
7.4 通信协议的测试 55
7.5 小结 56
第8章 总结与展望 57
8.1 总结 57
8.2 展望 57
致 谢 58
参考文献 59