基于SVM的高速公路路基病害自动检测算法  

Automatic Detection Algorithm for Expressway Subgrade Diseases Based on SVM

作　　者：周辉林[1] 姜玉玲[1] 徐立红 梁国卿 

机构地区：[1]南昌大学信息工程学院,江西南昌330031 [2]江西省天驰高速科技发展有限公司,江西南昌330103

出　　处：《中国公路学报》2013年第2期42-47,共6页China Journal of Highway and Transport

基　　金：国家自然科学基金项目(61062009);江西省科技支撑计划项目(2009BGB02200);江西省交通运输厅科技项目(2010H0017)

摘　　要：针对当前探地雷达(GPR)数据解释主要依赖专家经验存在的解释结果主观性强和数据解释周期长等问题,利用高速公路路基病害将导致其厚度和层界面反射信号的幅度发生改变等客观信息,结合探地雷达杂波抑制、层界面检测和平滑、感兴趣区域(ROI)提取、特征提取和模式识别技术,提出了一种新颖的高速公路路基病害自动检测算法,并利用该算法对江西省昌九高速公路南昌段采集的GPR数据进行了分析。研究结果表明:该算法的检测结果与结合专家经验和钻孔取芯样本构建的Ground Truth数据库的吻合度高达92.7%,且具有自动、快速等优越性,可为指导制定合适的养护策略及合理分配养护资金提供科学依据。Nowadays the analysis of ground penetrating radar（GPR） data mainly relies on the experts＇ experience, which may result in a series of problems such as subjective results and relatively long period of data interpretation. To solve these problems, a novel automatic detection algorithm for expressway subgrade diseases was proposed by using the information that subgrade diseases of expressway will lead to some changes of the thickness of pavement and the amplitude of the reflected signals from layer interfaces, and with GPR clutter suppression, level layer interface detection and smoothing, region of interest （ROI） extraction, feature extraction and pattern recognition technology. The GPR data collected from Nanchang section of Jiangxi Changjiu Expressway were analyzed. The results show that the agreement between the results of the algorithm proposed and the data from the Ground Truth database established with experts＇ experience and borehole coring is about 92.7~. This algorithm, an automatic and fast method, can provide scientific basis for formulating the suitable maintenance strategy and allocating the maintenance funds reasonably.

关 键 词：道路工程 高速公路 支持向量机 路基病害 自动检测 探地雷达 

分 类 号：U418.5[交通运输工程—道路与铁道工程]