Adaptive Prior Scene-Object SLAM for Dynamic Environments

2025-07-29

AI Summary

Traditional visual SLAM suffers from pose drift in dynamic environments due to its inherent static-scene assumption, especially under abrupt viewpoint changes and ambiguous features from moving objects. To address this, we propose a scene-object collaborative reliability assessment framework. Leveraging RGB-D data, our method jointly evaluates frame quality via scene-level geometric consistency and object-level motion discrimination, establishing a robust reference-frame selection mechanism. Furthermore, we introduce a prior-informed pose refinement strategy to enable dynamic pose correction and error suppression. Experiments on the TUM RGB-D dataset demonstrate that our approach significantly reduces absolute trajectory error (ATE) in dynamic scenes, achieving an average accuracy improvement of 32.7% over baseline methods. It maintains stable performance under severe viewpoint variations and high dynamic object density, outperforming state-of-the-art systems including ORB-SLAM2 and DynSLAM.

Technology Category

• Intelligent Robots: State Estimation
• Computer Vision: Motion & Tracking
• Planning, Routing, and Scheduling: Replanning and Plan Repair

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization

Abstract

Visual Simultaneous Localization and Mapping (SLAM) plays a vital role in real-time localization for autonomous systems. However, traditional SLAM methods, which assume a static environment, often suffer from significant localization drift in dynamic scenarios. While recent advancements have improved SLAM performance in such environments, these systems still struggle with localization drift, particularly due to abrupt viewpoint changes and poorly characterized moving objects. In this paper, we propose a novel scene-object-based reliability assessment framework that comprehensively evaluates SLAM stability through both current frame quality metrics and scene changes relative to reliable reference frames. Furthermore, to tackle the lack of error correction mechanisms in existing systems when pose estimation becomes unreliable, we employ a pose refinement strategy that leverages information from reliable frames to optimize camera pose estimation, effectively mitigating the adverse effects of dynamic interference. Extensive experiments on the TUM RGB-D datasets demonstrate that our approach achieves substantial improvements in localization accuracy and system robustness under challenging dynamic scenarios.

Problem

• Improves SLAM accuracy in dynamic environments
• Reduces localization drift from moving objects
• Enhances robustness against abrupt viewpoint changes

Innovation

• Scene-object reliability assessment framework
• Pose refinement using reliable frames
• Improved SLAM accuracy in dynamics