Research on Key Technologies of Detection and Recognition of

发布时间：2021-10-14 17:14

　　车辆的诞生不仅拉近了地区与地区之间的距离,而且加速了人类社会的发展。现今社会,人们的生活以及与车辆紧紧联系在一起,车辆已经成为人类生活中不可或缺的一部分。人们享受着汽车为我们带来的便利,但是不可忽视的是.随着道路上的车辆越来越多,也造成了许许多多的交通问题,如交通拥堵,交通事故和交通污染等,这些问题时刻威胁着我们的生活。为了解决这些日益增加的交通问题,仅仅依靠传统的交通治理方法并不能达到很好的效果,人们需要将交通治理与先进的高新技术结合起来,才能适应现代社会的发展,因此智能交通系统（ITS）便应运而生。智能交通系统融合了各种现代化的高新技术,将人类从繁杂的交通管理任务中解脱出来,实现了大范围的智能化、自动化的交通管理。智能交通管路系统可以搭载在交通管理部门,道路设施或者行人车辆上,辅助车辆驾驶,进行道路交通的引导与管理,但是由于技术局限,目前人们依然无法研究出能够大规模应用的智能交通系统。在日常的交通管理过程中,交通标志承载了最多的交通信息。驾驶员可以直观的从这些交通标志中获取前方道路信息与交通信息,交通部门也可以利用这些交通标志对道路交通进行方便、直接的管理。因此,当今的智能交通系统... 

【文章来源】：华中师范大学湖北省 211工程院校 教育部直属院校

【文章页数】：79 页

【学位级别】：硕士

【文章目录】：
Abstract
1 Introduction
    1.1 Background and research significance
    1.2 Research status at home and abroad
        1.2.1 Status of research on traffic sign detection
        1.2.2 Status of research on traffic sign recognition
    1.3 Difficulties in the recognition of traffic signs in real roads
    1.4 Main work in this paper
    1.5 Paper organization structure
2 Traffic Sign Detection and Recognition Technology
    2.1 China road traffic sign
    2.2 Traffic sign detection technology
        2.2.1 Color-based traffic sign detection
        2.2.2 Shape-based traffic sign detection
    2.3 Traffic sign recognition
        2.3.1 SVM
        2.3.2 Recognition method based on CNN
    2.4 Traffic sign recognition system architecture
    2.5 Summary of this chapter
3 Detection of Traffic Signs in Real Roads
    3.1 Preprocessing of traffic sign image of real road
        3.1.1 Gamma calibration
        3.1.2 RGB contrast enhancement algorithm
        3.1.3 Adaptive gamma calibration algorithm
    3.2 Color segmentation of traffic signs
        3.2.1 Red and blue color segmentation based on RGB color space
        3.2.2 Red, yellow and blue color separation based on HSV color space
        3.2.3 Combined color segmentation
    3.3 Shape-based traffic sign detection
        3.3.1 Morphological processing
        3.3.2 Shape-based traffic sign detection
    3.4 Overall process of traffic sign detection
    3.5 Summary of this chapter
4 Construction of Traffic Sign Recognition System in Real Roads
    4.1 Establishment and expansion of data sets
    4.2 Traffic sign recognition method based on HOG feature and SVM
        4.2.1 HOG feature extraction
        4.2.2 SVM training
    4.3 Based on improved AlexNet traffic sign recognition method
        4.3.1 Classic AlexNet network
        4.3.2 Recognition of traffic signs based on improved AlexNet
    4.4 Construction of traffic sign recognition system
        4.4.1 MATLAB GUI
        4.4.2 Traffic sign recognition system
    4.5 Traffic sign recognition system performance evaluation
        4.5.1 Campus road traffic sign recognition test
        4.5.2 Off-campus road scene traffic sign recognition test
        4.5.3 Traffic sign recognition test for different resolution images
    4.6 Summary of this chapter
5 Summary and Outlook
    5.1 Summary
    5.2 Deficiencies in existence and prospects for future research work
References
Acknowledgements
Appendix A
    Chinese abstract


【参考文献】：
期刊论文
[1]最优RGB线性组合颜色模型目标检测方法[J]. 温芝元,曹乐平.  计算机工程与应用. 2015(18)
[2]基于网格搜索的PCA-SVM道路交通标志识别[J]. 吴峰,陈后金,姚畅,郝晓莉.  铁道学报. 2014(11)



本文编号：3436530