Normal-guided Detail-Preserving Neural Implicit Function for High-Fidelity 3D Surface Reconstruction

TL;DR

This paper introduces a neural implicit method that leverages surface normals derived from depth maps to improve high-fidelity 3D surface reconstruction from sparse multi-view RGB images, capturing fine details and thin structures.

Contribution

It proposes a novel normal-guided training approach for neural implicit functions that enhances detail preservation in 3D reconstruction from limited views.

Findings

Achieves state-of-the-art accuracy with minimal views

Successfully captures intricate geometric details

Effective in reconstructing thin structures

Abstract

Neural implicit representations have emerged as a powerful paradigm for 3D reconstruction. However, despite their success, existing methods fail to capture fine geometric details and thin structures, especially in scenarios where only sparse multi-view RGB images of the objects of interest are available. This paper shows that training neural representations with first-order differential properties (surface normals) leads to highly accurate 3D surface reconstruction, even with as few as two RGB images. Using input RGB images, we compute approximate ground-truth surface normals from depth maps produced by an off-the-shelf monocular depth estimator. During training, we directly locate the surface point of the SDF network and supervise its normal with the one estimated from the depth map. Extensive experiments demonstrate that our method achieves state-of-the-art reconstruction accuracy with a minimal number of views, capturing intricate geometric details and thin structures that were previously challenging to capture.