A Multi-View Vision System for Astronaut Postural Reconstruction with Self-Calibration

• Published in: Aerospace 2023-03, Vol.10 (3), p.298
• Main Authors: Gan, Shuwei, Zhang, Xiaohu, Zhuge, Sheng, Ning, Chenghao, Zhong, Lijun, Li, You
• Format: Article
• Language: English
• Subjects: Accuracy; Astronaut performance; astronaut performance capture; Astronauts; Calibration; Cameras; Discovery and exploration; extrinsic camera calibration; Human body; Human performance; human pose estimation; Kinematics; Modules; Moon; Motion capture; multi-view system; Neural networks; Optimization; Outer space; Parameter estimation; Parameters; Posture; Reconstruction; Self calibration; Sensors; Space exploration; Space missions; Space stations; Spacecraft modules; Vision systems; Workloads
• ISSN: 2226-4310; 2226-4310
• DOI: 10.3390/aerospace10030298

Space exploration missions involve significant participation from astronauts. Therefore, it is of great practical importance to assess the astronauts’ performance via various parameters in the cramped and weightless space station. In this paper, we proposed a calibration-free multi-view vision system for astronaut performance capture, including two modules: (1) an alternating iterative optimization of the camera pose and human pose is implemented to calibrate the extrinsic camera parameters with detected 2D keypoints. (2) Scale factors are restricted by the limb length to recover the real-world scale and the shape parameters are refined for subsequent postural reconstruction. These two modules can provide effective and efficient motion capture in a weightless space station. Extensive experiments using public datasets and the ground verification test data demonstrated the accuracy of the estimated camera pose and the effectiveness of the reconstructed human pose.