A point cloud optimization method for accurate line-scan 3-D measurement in motion with dynamic disturbance

In-motion three-dimensional (3-D) shape measurement with dense point cloud is widely required, and line-scan cameras are promising with the advantage of high-frequency high-resolution measurement. However, line-scan camera measurement lacks constraints in point cloud stitching direction and dynamic disturbance in motion distorts the stitched point cloud if there's no precise guide rail. Point cloud optimization remains a challenge and the applications of line-scan cameras hit a bottleneck. In this work, accurate line-scan 3-D measurement is achieved despite dynamic disturbance. Instantaneous two-dimensional imaging of an attached area-scan camera provides constraints for line-scan point cloud profiles, and algorithms are proposed for effective matching and optimization. Local continuity constraints are further utilized for comprehensive line-scan point cloud optimization. Through quantitative and qualitative evaluations, the reconstruction accuracy is significantly improved in motion with dynamic disturbance. The proposed method promotes the application of line-scan cameras and solves a significant challenge in 3-D shape measurement in motion.