Multishot Structured-Light 3-D Scanning for Surfaces in Challenging Motion

Minghui Duan,Yabing Zheng,Yi Jin,Jinjin Zheng,Enhong Chen

DOI: https://doi.org/10.1109/tie.2024.3447752

IF: 7.7

2024-01-01

IEEE Transactions on Industrial Electronics

Abstract:Challenging motion, resulting in serious motion artifacts, is a well-known problem in structured-light (SL) 3-D scanning. Single-shot imaging or tracking interframe offsets provides a solution; however, these methods either sacrifice spatial resolution or work on slow motion under quasi-still photography. This article presents a universal multishot SL (MS-SL) strategy to measure surfaces in challenging motion. During the acquisition of MS-SL 3-D scanning, serious motion artifacts diffuse into the overall image sequence, showing long-range and short-range effects. To quantify the different effects of motion artifacts, we establish a spatiotemporal artifact coupling (STAC) model fusing the interframe ghosts and intraframe motion blur. Furthermore, without complicated decoupling, a unified deghosting and deblurring (UDD) framework is constructed for automatically removing the coupled motion artifacts. Unlike existing restoration methods functioning on natural images, UDD can recover the intensity and phase distributions under periodically encoded illumination. Through the adaptive extraction and utilization of spatiotemporal mapping, our strategy can be applied to various SL 3-D scanning technologies. Simulations and experiments prove that our strategy extends the excellent metrological performance of SL 3-D scanning to various challenging motion scenes and achieves a measurement accuracy of 0.03 mm.