TrCPDs: Trimmed coherent point drift methods for robust fine registration of low-overlap point cloud

Key Points

• Robust registration improves the alignment of low-overlap point clouds, enhancing geometric accuracy.
• Key evidence shows a significant reduction in alignment error by up to 30% compared to traditional methods.
• This analysis assesses trimmed coherent point drift methods across various datasets, focusing on noise reduction.
• Such findings highlight the methods' potential for accurate point cloud registration in real-world applications.