Highly Accurate Extrinsic Calibration of Terrestrial Laser Scanner and Digital Camera for Structural Monitoring Applications

In the current state-of-the-art, monitoring and analysis of shortand long-term deformations of natural and artificial objects (e.g. dams, bridges, towers, landslides,...) has received increasing interest by researchers and operators who are involved in this field of Geodetic science and engineering. In order to improve the accuracy of deformation monitoring and analysis, terrestrial laser scanner (TLS) and high resolution digital camera are integrated to optimally combine their information. TLS can directly provide 3D spatial coordinates of the object surface in combination with reflectivity values. In addition, digital cameras can acquire high resolution images. The information from both camera and TLS are complementary to each other. The strength of one measurement method overcomes the weakness of the other one. In order to fuse the acquired data from these two kinds of sensors in the common coordinate system, it is a vital and pre-requisite step to precisely determine the extrinsic parameters between TLS and digital camera. Thus, the calibration of the aforementioned hybrid sensor system can be separated into three single calibrations: calibration of the camera, calibration of the TLS and extrinsic calibration between TLS and digital camera. Camera calibration and TLS calibration (internal error sources) are commonly performed in advance in a 3D laboratory. In this research, we focus on the high accurate estimation of extrinsic parameters between fused sensors on the basis of target-based calibration using different types of observations (image measurements, TLS measurements and laser tracker measurements for validation) by considering different weights. Space resection bundle adjustment is solved based upon Gauss-Markov and Gauss-Helmert model. At the end, three case studies are investigated in the 3D laboratory and numerical results of the aforementioned calibration are presented and discussed. The results depict that high accurate extrinsic parameters are achieved with applied Gauss-Helmert model and variance component estimation in the integrated multisensor-system. Finally, a real world case study of a selected structural monitoring project is presented and discussed.