Abstract:
Objectives High density 3D laser scanning point cloud contains massive data, leading to low efficiency in detection of tunnel over-under break. In order to get the results while measuring, it is necessary to solve problems including fast extraction of tunnel centerline, point cloud data down-sampling and efficient over-under break calculation.
Methods First, we propose a centerline extraction algorithm based on the outer rectangle of tunnel plane and the design file, which can construct the centerline quickly and effectively. Then, we explore the efficiency improvement of section extraction and over-under break calculation by setting the optimal down-sampling parameter. Finally, we propose an over-under break calculation algorithm based on inner contour using inverse distance weighted method.
Results Experiment shows that the proposed algorithm can realize online calculation, generate reports and construct 3D over-under break model. It takes 247.14 s to get over-under break calculation results using the georeferenced point cloud data of 3 measurement stations. Compared with the measured values by total station, the mean square error of the over-under break results of a 40 m section is only 0.96 mm.
Conclusions The proposed algorithm can detect the over-under break in tunnel constructing and meet the requirements of high-precision and immediateness, which could be used during construction of the tunnel project.
下载:
