From Open Waters to Enclosed Cabins: ProteusVPR for Cross-Scene Visual Place Recognition in Maritime Perception and Cabin Inspection

2026-06-23

AI Summary

This study addresses the challenge of cross-scene visual place recognition (VPR) in maritime environments, where significant visual discrepancies exist between open decks and enclosed cabins. To tackle this problem, the authors propose ProteusVPR, a two-stage retrieval-and-refinement framework: the first stage employs a standard VPR model for initial candidate retrieval, while the second stage fuses information from the current and preceding two frames, integrating geometric descriptors, local affine coordinate systems, and camera orientation encoding to achieve precise localization. This work presents the first VPR method specifically designed for maritime cross-scene scenarios and introduces XHZ, the first 8K panoramic shipboard dataset, to enable rigorous evaluation. Experiments demonstrate that ProteusVPR reduces average localization error by over 60% on XHZ compared to existing methods, confirming its effectiveness and robustness in complex maritime settings.

Abstract

Autonomous robotic inspection in maritime environments presents unique challenges for Visual Place Recognition (VPR) due to cross-scene perceptual shifts. Robots navigating ship-borne environments must transition between visually distinct domains: open decks with sparse textures and severe illumination changes, and enclosed cabins with repetitive structures and high visual ambiguity. Existing VPR methods, designed primarily for urban or indoor scenes, fail to generalize reliably across these starkly different scenarios. To address this, we propose ProteusVPR, a two-stage retrieval-refinement framework. The first stage employs any standard VPR model for initial image retrieval. The second stage introduces a geometric-visual estimation network that fuses the retrieved image with two temporally preceding frames, incorporating geometric descriptors, a local affine coordinate system, and camera azimuth encoding to achieve precise localization. To support this task, we introduce the XHZ dataset, an 8K-panoramic ship-borne dataset collected from an operational vessel, featuring multi-floor cabin structures, deck transition zones, and strict query-database separation for rigorous evaluation. Extensive experiments on the XHZ dataset demonstrate that ProteusVPR consistently improves the localization accuracy across multiple VPR backbones, reducing mean localization error by over 60\% on average and that ProteusVPR offers an effective and robust solution for precise visual localization in challenging, cross-scene maritime environments.

Problem

• Visual Place Recognition
• Cross-Scene Perception
• Maritime Robotics
• Autonomous Inspection
• Perceptual Shift

Innovation

• Visual Place Recognition
• Cross-Scene Localization
• Geometric-Visual Fusion
• Maritime Robotics
• Two-Stage VPR Framework