This study aims to address wear, cracks and material degradation in critical components within aerospace, automotive and energy industries. It focuses on developing an efficient and precise trajectory planning method for repairing and reinforcing worn surfaces on complex curved parts, using three-dimensional point cloud reconstruction based on binocular vision technology.
Binocular cameras capture damaged surface images, and precise three-dimensional point clouds of the failed surface are generated through stereo matching and three-dimensional reconstruction. A novel method creates reinforcement trajectories aligned with the point cloud, ensuring the deposition head’s axis parallels the point cloud normal. This approach bypasses STereoLithography (STL) or non-uniform rational B-spline (NURBS) surface generation, directly processing point cloud data.
Experimental results demonstrate that the proposed method outperforms traditional STL model-based approaches, achieving a 33% improvement in trajectory position accuracy and reducing planning time by approximately 30%. The method ensures precise positioning and orientation of the cladding head, facilitating continuous and uniform material deposition over complex curved surfaces. This results in high-quality cladding of damaged areas, significantly enhancing both work efficiency and repair performance.
This study introduces an innovative trajectory planning method that directly processes point cloud data for repairing complex curved surfaces, eliminating intermediate steps like STL or NURBS surface generation. The approach enhances trajectory accuracy, planning efficiency and deposition head orientation control, ensuring superior repair quality. It provides a practical solution for extending the service life of critical components in high-stakes industries, advancing repair and reinforcement technologies.
