High precision slope deformation monitoring by UAV with industrial photogrammetry

Abstract

As an important basis for quality and safety assessment during the service of large-scale structures, deformation monitoring has received extensive attention in both civil and mechanical engineering. Although many methods of deformation monitoring have been developed, measurement accuracy, implementation cost, and convenience are still the critical barriers that prevent their wide application, especially for side slopes measuring thousands of kilometres. This work proposes a high-accuracy, low-cost, and automatic slope deformation monitoring method that can be used with unmanned aerial vehicles (UAVs) and industrial photogrammetry technology. Specifically, a UAV is equipped with a high-quality industrial camera, the slope surface is arranged with artificial markers, and five markers in the stable area near the bottom of the slope are pre-calibrated to define a local 3D coordinate system. The flight of the UAV, detection of the markers, and determination of their 3D coordinates are all completed automatically. To test this method, a monitoring experiment of the cross section of the slope of the Chinese South-to-North Water Diversion Project (CSNWDP) was performed. It is shown that this method improves measurement accuracy by an order of magnitude over conventional aerial photogrammetry, and both the horizontal and vertical displacements are measured accurately within 2~3 mm.