Development of express assessment system of road surface with stereo image processing and GPS

本研究开发了一套用于路面评价的立体摄像系统。在这项研究中，首先利用校准的立体摄像系统拍摄的静态照片来重建路面中目标的三维坐标。在此基础上，给出了合适的初始条件和图像相关方法，提出了一种改进的迭代Closet Point方法，以对不同视点网格得到的不同坐标进行对齐。为了评估该系统的性能，已经进行了几次现场测试。该系统在成功对准所有测量坐标后，不仅可以提供路面局部缺陷的准确信息，如修补、裂缝、坑洞等，还可以提供长距离路面全局波动的准确信息。

This research developed a stereo camera system for road surface assessment. In this study, first the static photos taken by a calibrated stereo camera system are utilized to reconstruct the three-dimensional coordinates of targets in the pavement. Subsequently to align the various coordinates obtained from different view meshes, one modified Iterative Closet Point method is proposed by affording the appropriate initial conditions and image correlation method. Several field tests have been carried out to evaluate the capabilities of this system. After succeeding to align all the measured coordinates, this system can offer not only the accurate information of local deficiency such as the patching, crack or pothole, but also global fluctuation in a long distance range of the road surface.