Geometric Constrained Non-Line-of-Sight Imaging

2025-03-23

AI Summary

In non-line-of-sight (NLOS) imaging, jointly reconstructing surface normals and albedo of hidden objects is critical for improving geometric accuracy and scene understanding; however, this joint estimation elevates the problem to a high-dimensional tensor optimization with prohibitive computational cost. To address this, we propose the first normal-field regularization framework for NLOS surface reconstruction, introducing a geometric constraint based on the Frobenius norm of the shape operator to enable efficient joint optimization of normals and albedo. Our method integrates transient light transport modeling with intrinsic geometric priors, achieving superior geometric reconstruction accuracy—on transient data acquired in just 15 seconds—while accelerating runtime by 30× over state-of-the-art methods. It demonstrates strong robustness on both synthetic and real-world measurements. The core contribution lies in transcending conventional matrix-based functional modeling by establishing an efficient, differentiable geometric regularization framework within tensor space.

Technology Category

• Search and Optimization: Non-convex Optimization
• Computer Vision: Learning & Optimization for CV
• Machine Learning: Matrix & Tensor Methods

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Security and Privacy: Data transparency and provenance
• Responsible Web: Algorithmic accountability and transparency on the web

Abstract

Normal reconstruction is crucial in non-line-of-sight (NLOS) imaging, as it provides key geometric and lighting information about hidden objects, which significantly improves reconstruction accuracy and scene understanding. However, jointly estimating normals and albedo expands the problem from matrix-valued functions to tensor-valued functions that substantially increasing complexity and computational difficulty. In this paper, we propose a novel joint albedo-surface reconstruction method, which utilizes the Frobenius norm of the shape operator to control the variation rate of the normal field. It is the first attempt to apply regularization methods to the reconstruction of surface normals for hidden objects. By improving the accuracy of the normal field, it enhances detail representation and achieves high-precision reconstruction of hidden object geometry. The proposed method demonstrates robustness and effectiveness on both synthetic and experimental datasets. On transient data captured within 15 seconds, our surface normal-regularized reconstruction model produces more accurate surfaces than recently proposed methods and is 30 times faster than the existing surface reconstruction approach.

Problem

• Jointly estimating normals and albedo increases complexity in NLOS imaging
• Improving normal field accuracy enhances hidden object detail representation
• Achieving high-precision reconstruction of hidden object geometry efficiently

Innovation

• Uses Frobenius norm for normal field control
• First regularization method for hidden normals
• 30 times faster than existing reconstruction methods