Abstract: 3D laser scanning technology has been widely used in tunnel deformation detection, but there are some problems such as large amount of point cloud data, low efficiency in section extraction, and poor universality of different types of tunnel deformation detection algorithms. This article proposes a fast search algorithm for cutting local range point clouds of tunnel cross-sections,an octree index is established for point cloud data, and in order to reduce the complexity of the algorithm and simplify the search process, the query point coordinates, cutting thickness, and offset are given to define the range plane polygon of the search area, and it performs intersection detection with the node polygon of octree index to quickly locate and extract tunnel cross-sections.Divide the contour line of the tunnel design section equally and complete the numerical calculation of deformation of any shaped tunnel based on laser point cloud measurement. Finally, taking the Daliang tunnel of the Lanxin Passenger Dedicated Line affected by the earthquake as an example, the setting accuracy of scanning station is statistically analyzed,the extraction of tunnel sections and deformation detection analysis are completed, and the offset patterns of different mileage intervals are provided, providing data support for post earthquake repair design.

Key words: section extraction, octree index, intersection of planar polygons, projection transformation, divide the design section equally, deformation detection of post earthquake