An integrated calibration technique for variable-boresight three-dimensional imaging system

• Published in: Optics and lasers in engineering Vol. 153; p. 107005
• Main Authors: Liu, Xingsheng, Li, Anhu
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.06.2022
• Subjects: Flexible reference transfer; Integrated model calibration; Multi-parameter identification; Three-dimensional reconstruction; Variable-boresight imaging; Multi-parameter identification; PCS; CCD; Variable-boresight imaging; Three-dimensional reconstruction; FRT; RMSE; FOV; RPM; Integrated model calibration; 3D / 2D; PnP; Flexible reference transfer; AOE
• ISSN: 0143-8166; 1873-0302
• DOI: 10.1016/j.optlaseng.2022.107005

•Unified self-calibration framework for wedge-prism-based 3D imaging system.•Multi-parameter integrated calibration based on flexible reference transfer.•Explicit model representation for misalignment between camera and wedge prism. There is an emerging need for multiview stereo systems achieved with high accuracy, wide coverage, and flexible perspective in spatially-limited vision and close-range photogrammetry. Accurate and efficient calibration is a critical prerequisite that is always performed to ensure high-performance object reconstruction and situational awareness. This paper proposes an integrated calibration technique based on rigorous modeling for the variable-boresight three-dimensional imaging system composed of a single camera and a wedge prism. A unified self-calibration framework which considers the misalignment between camera and prism is systematically demonstrated such that all involved parameters are estimated by minimizing the statistical discrepancy between model prediction and reference information. Moreover, a novel approach is presented to construct calibration reference through adaptive generation and flexible transfer. Experiments have validated that the proposed technique can offer a viable solution to multi-parameter identification for the imaging system, which therefore contributes to the recovery of any spatial object with competitive geometric accuracy and satisfactory robustness to potential error sources.