路面三维病害自动检测技术研究

Research on Automatic Detection Technology of Three Dimensional Road Surface Defects

对于路面沉陷、坑槽等三维病害,人工现场检测主要通过3 m直尺法来进行测量。而如何对道路综合检测车拍摄的三维图像进行测量,规范中并没有明确的规定。为解决这一问题,提出了一种模拟3 m直尺的分析计算模型,将三维图像转换成二维的高差图,再在二维的高差图上抽出路框差、错台、波浪、坑槽、沉陷、车辙等三维病害的位置、尺寸及严重程度,并将该计算模型用于实际试验的数据分析,取得了预期的结果。这一计算模型实现了三维病害自动化检测和人工检测标准的统一,并为三维病害的计算机自动化分析铺平了道路。

For three-dimensional defects such as road subsidence and potholes,manual on-site inspection is mainly measured using the 3-meter straightedge method.There is no clear regulation in the specifications on how to measure the three-dimensional images captured by road comprehensive inspection vehicles.To solve this problem,a simulated 3-meter straightedge analysis and calculation model is proposed,which transforms the three-dimensional images into two-dimensional elevation maps.Then,the positions,dimensions,and severity of three-dimensional diseases such as road frame differences,misalignment,waves,potholes,subsidence,and ruts are extracted from the two-dimensional elevation map.The calculation model is used for data analysis in actual experiments and the expected results are achieved.This computational model harmonizes automated and manual detection standards for three-dimensional defects,and paves the way for computer automated analysis of three-dimensional defects.