A Self-Recalibrated 3D Vision System for Accurate 3D Tracking in Hypersonic Wind Tunnel

3D vision system becomes a critical tool in wind tunnel experiment. However, in hypersonic wind tunnel vibration is unavoidable, it will disturb the pose of cameras in vision system and heavily decrease the measurement accuracy. Aiming at solving this problem, an accurate and stable two-step self-calibration method is proposed. Firstly, a stable and fast non-iterative method is adopted to calculate the initial value of camera pose. Then, a globally convergent orthogonal iterative algorithm is used to obtain an accurate solution. According to the experiment results, the 3D vision system with self-calibration method can get accurate measurement results in hypersonic wind tunnel. And we applied the 3D vision system to measure the deformation of an aircraft wing model in hypersonic wind tunnel of China Aerodynamics Research and Development Center successfully.