SGCR: Spherical Gaussians for Efficient 3D Curve Reconstruction

TL;DR

This paper introduces Spherical Gaussians, a new 3D geometric boundary representation optimized through view-based rendering loss, enabling efficient and accurate 3D curve reconstruction from multi-view images, surpassing existing methods.

Contribution

The paper proposes Spherical Gaussians for 3D boundary representation and SGCR algorithm for direct parametric curve extraction, improving accuracy and efficiency in 3D edge reconstruction.

Findings

SGCR outperforms state-of-the-art in 3D edge reconstruction.

Spherical Gaussians enable direct 3D curve reconstruction without 3D supervision.

The method is efficient and effective for multi-view 3D reconstruction tasks.

Abstract

Neural rendering techniques have made substantial progress in generating photo-realistic 3D scenes. The latest 3D Gaussian Splatting technique has achieved high quality novel view synthesis as well as fast rendering speed. However, 3D Gaussians lack proficiency in defining accurate 3D geometric structures despite their explicit primitive representations. This is due to the fact that Gaussian's attributes are primarily tailored and fine-tuned for rendering diverse 2D images by their anisotropic nature. To pave the way for efficient 3D reconstruction, we present Spherical Gaussians, a simple and effective representation for 3D geometric boundaries, from which we can directly reconstruct 3D feature curves from a set of calibrated multi-view images. Spherical Gaussians is optimized from grid initialization with a view-based rendering loss, where a 2D edge map is rendered at a specific view and then compared to the ground-truth edge map extracted from the corresponding image, without the need for any 3D guidance or supervision. Given Spherical Gaussians serve as intermedia for the robust edge representation, we further introduce a novel optimization-based algorithm called SGCR to directly extract accurate parametric curves from aligned Spherical Gaussians. We demonstrate that SGCR outperforms existing state-of-the-art methods in 3D edge reconstruction while enjoying great efficiency.