Computer vision application programming for settlement monitoring in a drainage tunnel

Abstract

The study employs computer vision technology to pose a new type of optical instrument composited of the Raspberry Pi, digital cameras and chessboards for structural settlement monitoring. Using the OpenCV functions findChessboardCorners and cornerSubPix, displacement measurements in five distances between a camera and chessboard at 15, 17, 20, 23 and 25 m were undertaken to analyze measuring standard deviations that were from 0.027 to 0.048 pixels in laboratory tests. For field testing, five optical settlement instruments were installed at a drainage tunnel that was constructed at an 80-m-depth location within a landslide area. The resolution and accuracy of the instrument can be determined at 0.01 and 0.11 cm, respectively, as an optimally installed distance of 20 m in the drainage tunnel. The maximum settlement amount of field monitoring was 0.61 cm in six months. Overall, the optical instrument is more cost-effective and IOT-based for a long-term settlement monitoring.