基于路面高精三维图像的车辙识别与评价
发布时间：2019-02-19 17:46
【摘要】：沥青路面作为我国高等级公路路面的主要形式,其服务性能的好坏直接影响到了人们出行的舒适感,交通的顺畅性。随着越来越多的高等级沥青路面投入使用,如何对已建道路路面进行实时的检测、养护管理,成了道路管理者需要面临的一大问题。而科学合理的对路面破损状况进行检测和评价是进行养护管理的重要前提。本文主要以车辙破损为研究对象,从车辙的检测设备、检测方法、车辙图像处理、车辙特征提取以及车辙指标建立这几个方面展开了研究。首先介绍路面破损的主要类型,针对车辙破损的成因,危害进行主要描述,并提出现有车辙检测系统和评价方法的不足和存在的问题。随后对公路路面检测系统和设备的多样性进行简要描述,对于车辙病害的检测方法进行主要介绍,分别列出各检测设备的优缺点。介绍DHDV数字化公路检测车的检测原理和激光三维扫描系统以及高精度路面三维图像与二维图像的不同之处。第三部分是对路面高精三维图像的预处理。借鉴已有的二维图像的去噪处理方法,提出适用于高精度三维图像的车辙病害的去噪算法。并对去噪处理后的路面横断面曲线与人工绘制的横截面曲线进行对比,验证算法的有效性以及稳定性。最后是车辙特征的提取和车辙指标的建立。根据第四章获取的路面横截面车辙曲线,提取车辙的深度、面积、宽度等数据,建立了能够反映车辙破损状况的车辙评价指标。
[Abstract]:Asphalt pavement is the main form of high grade highway pavement in our country. The service performance of asphalt pavement has a direct impact on people's travel comfort and traffic smoothness. With more and more high-grade asphalt pavement being put into use, how to detect and maintain the pavement in real time has become a big problem that road managers need to face. The scientific and reasonable detection and evaluation of pavement damage is an important prerequisite for maintenance management. In this paper, rut damage is the main research object, from the rut detection equipment, detection methods, rut image processing, rut feature extraction and rut index establishment, the paper studies the rut detection equipment, detection method, rut image processing, rut feature extraction and rut index establishment. This paper first introduces the main types of pavement damage, mainly describes the causes and hazards of rutting damage, and puts forward the shortcomings and problems of the existing rut detection system and evaluation methods. Then the diversity of highway pavement detection system and equipment is briefly described, and the detection methods of rutting diseases are mainly introduced, and the advantages and disadvantages of each detection equipment are listed respectively. This paper introduces the detection principle of DHDV digital highway inspection vehicle, the laser three-dimensional scanning system and the difference between high-precision three-dimensional image and two-dimensional image of road surface. The third part is the preprocessing of high-precision three-dimensional road surface image. Based on the existing two-dimensional image denoising methods, a rut disease denoising algorithm for high-precision 3D images is proposed. The effectiveness and stability of the algorithm are verified by comparing the cross-section curve of road surface after denoising with that of manual drawing. Finally, the extraction of rutting feature and the establishment of rutting index. According to the rutting curve of road cross-section obtained in the fourth chapter, the data of rut depth, area and width are extracted, and the rut evaluation index which can reflect the rut damage is established.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U418.6
本文编号：2426719
[Abstract]:Asphalt pavement is the main form of high grade highway pavement in our country. The service performance of asphalt pavement has a direct impact on people's travel comfort and traffic smoothness. With more and more high-grade asphalt pavement being put into use, how to detect and maintain the pavement in real time has become a big problem that road managers need to face. The scientific and reasonable detection and evaluation of pavement damage is an important prerequisite for maintenance management. In this paper, rut damage is the main research object, from the rut detection equipment, detection methods, rut image processing, rut feature extraction and rut index establishment, the paper studies the rut detection equipment, detection method, rut image processing, rut feature extraction and rut index establishment. This paper first introduces the main types of pavement damage, mainly describes the causes and hazards of rutting damage, and puts forward the shortcomings and problems of the existing rut detection system and evaluation methods. Then the diversity of highway pavement detection system and equipment is briefly described, and the detection methods of rutting diseases are mainly introduced, and the advantages and disadvantages of each detection equipment are listed respectively. This paper introduces the detection principle of DHDV digital highway inspection vehicle, the laser three-dimensional scanning system and the difference between high-precision three-dimensional image and two-dimensional image of road surface. The third part is the preprocessing of high-precision three-dimensional road surface image. Based on the existing two-dimensional image denoising methods, a rut disease denoising algorithm for high-precision 3D images is proposed. The effectiveness and stability of the algorithm are verified by comparing the cross-section curve of road surface after denoising with that of manual drawing. Finally, the extraction of rutting feature and the establishment of rutting index. According to the rutting curve of road cross-section obtained in the fourth chapter, the data of rut depth, area and width are extracted, and the rut evaluation index which can reflect the rut damage is established.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U418.6
【参考文献】
相关期刊论文 前3条
1 周星;高志军;;立体视觉技术的应用与发展[J];工程图学学报;2010年04期
2 侯相深,王哲人,刘振鹏;路面损坏图像的自动采集与处理设备的技术探究[J];公路;2003年02期
3 侯相深;马松林;王彩霞;;基于行车安全的沥青路面车辙测量与评价指标的研究[J];公路交通科技;2006年08期
相关硕士学位论文 前3条
1 薛英;沥青路面车辙检测算法研究[D];南京理工大学;2011年
2 柳钰;沥青路面车辙深度检测系统研究[D];武汉理工大学;2007年
3 周杰;应用于图像处理的中值滤波改进算法[D];北京邮电大学;2007年
，本文编号：2426719
本文链接：https://www.wllwen.com/kejilunwen/daoluqiaoliang/2426719.html
