Single-View Rolling-Shutter SfM

2026-03-12

AI Summary

This work addresses the lack of a systematic solution for structure-from-motion (SfM) reconstruction from a single rolling-shutter (RS) camera view. It presents the first geometric model for single-view RS imaging, explicitly characterizing the motion and scene geometry constraints encoded by point and line features in the image. The study identifies the set of recoverable parameters and establishes a minimal problem formulation tailored to this setting. Through careful algebraic derivation and the design of customized solvers, the approach leverages both point and line correspondences to validate its feasibility. Extensive experiments across diverse representative scenes demonstrate the method’s effectiveness while also revealing fundamental theoretical and practical limitations inherent to single-view RS SfM.

Technology Category

• Computer Vision: Scene Analysis & Understanding
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Multi-instance/Multi-view Learning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Security and Privacy: Large-scale security measurements

Abstract

Rolling-shutter (RS) cameras are ubiquitous, but RS SfM (structure-from-motion) has not been fully solved yet. This work suggests an approach to remedy this: We characterize RS single-view geometry of observed world points or lines. Exploiting this geometry, we describe which motion and scene parameters can be recovered from a single RS image and systematically derive minimal reconstruction problems. We evaluate several representative cases with proof-of-concept solvers, highlighting both feasibility and practical limitations.

Problem

• Rolling-shutter
• Structure-from-Motion
• Single-view geometry
• Camera motion
• 3D reconstruction

Innovation

• Rolling-shutter
• Single-view geometry
• Structure-from-Motion
• Minimal solvers
• Camera motion recovery