Targetless approach for Automated Calibration of Multi-Sensor Systems

In this thesis, we investigate the feasibility of employing an adapted Structural Similarity Index Measurement (SSIM) methodology for both extrinsic and intrinsic calibration in a multi-sensor system composed of a camera and a LiDAR. The study begins by evaluating the impact of adjusting a single degree of freedom (DOF) on the resulting SSIM values. Subsequently, the analysis is extended to multiple DOFs. Experimental validation was performed across several datasets, including one acquired in the EOS laboratory. Although the findings indicate that SSIM-based calibration is potentially viable, numerous unresolved challenges remain, rendering the current calibration outcomes insufficiently accurate for mobile multi-sensor systems that traverse in 6 DOFs.