用于路面车辙检测的线结构光图像处理流程

Line-Structured Light Image Processing Procedure for Pavement Rut Detection

摘要:

激光线结构光可用于路面车辙检测。分析了激光线结构光车辙检测原理，提出了适应沥青路面纹理特征和技术状况的线结构光图像处理流程，包括路面横断面曲线提取和车辙特征参数提取。其中路面横断面曲线提取由光心提取、光心连接和光心曲线平滑实现。基于高斯拟合的光心提取用于确定线结构光位置，光心连接用线性插值得到完整的路面横断面曲线，而光心曲线平滑则可削弱路表纹理及其他干扰因素的影响。车辙特征参数提取在获取横断面包络线的基础上实现，最大车辙深度和曲线填充面积均可作为车辙评价参数，而路面横断面曲线也可用于路面横向变形特征的描述。模拟车辙试验表明，线结构光检测可以精确测量车辙深度，本文所述图像处理流程具备可行性。

Abstract:

Pavement images with the laser line-structured light pattern can be used to detect the pavement rut. The work principle of the laser line-structured light in rut detection is introduced in this study and a line-structured light image processing method is proposed to deal with the complicated characteristics of the pavement surface textures as well as the technical conditions. The image processing procedure can be divided into two parts, the extraction of the pavement cross-section curve and the output of the rutting parameters. The pavement transverse section curve can be obtained through three steps. First, light center extraction, which aims to find the true center of the line-structured light through Gaussian fitting. Second, light center connection, which restores the characteristics of the entire pavement transverse section by linear interpolation. Third, smoothing of the light center curve, which can reduce the impact of pavement surface texture and other interferences. The output of the rutting parameters is based on the envelope line of the pavement transverse section curve. The maximum rut depth and the rut area can be calculated as the rutting parameters, and the pavement transverse section curve can also be used to describe the characteristics of the pavement transverse deformation. The rut sample is used to test the rut measurement method and the image processing procedure above. The result shows that the line-structured light can measure the rut depth accurately and the processing procedure proposed in this work has the potential to be applicated in practice.