Diffusing Winding Gradients (DWG): A Parallel and Scalable Method for 3D Reconstruction from Unoriented Point Clouds

Abstract

This paper presents Diffusing Winding Gradients (DWG) for reconstructing watertight surfaces from unoriented point clouds. Our method exploits the alignment between the gradients of screened generalized winding number (GWN) field–a robust variant of the standard GWN field– and globally consistent normals to orient points. Starting with an unoriented point cloud, DWG initially assigns a random normal to each point. It computes the corresponding sGWN field and extract a level set whose iso-value is the average GWN values across all input points. The gradients of this level set are then utilized to update the point normals. This cycle of recomputing the sGWN field and updating point normals is repeated until the sGWN level sets stabilize and their gradients cease to change. Unlike conventional methods, DWG does not rely on solving linear systems or optimizing objective functions, which simplifies its implementation and enhances its suitability for efficient parallel execution. Experimental results demonstrate that DWG significantly outperforms existing methods in terms of runtime performance. For large-scale models with 10 to 20 million points, our CUDA implementation on an NVIDIA GTX 4090 GPU achieves speeds 30-120 times faster than iPSR, the leading sequential method, tested on a high-end PC with an Intel i9 CPU. Furthermore, by employing a screened variant of GWN, DWG demonstrates enhanced robustness against noise and outliers, and proves effective for models with thin structures and real-world inputs with overlapping and misaligned scans. For source code and additional results, visit our project webpage: https://dwgtech.github.io/.