GV-Bench: Benchmarking Local Feature Matching for Geometric Verification of Long-term Loop Closure Detection

TL;DR

This paper introduces GV-Bench, a comprehensive benchmark for evaluating geometric verification methods in long-term loop closure detection, addressing a critical gap in visual SLAM robustness.

Contribution

It proposes a unified benchmark for geometric verification in long-term visual localization and evaluates six local feature matching methods to analyze their limitations.

Findings

Benchmark reveals strengths and weaknesses of different feature matching methods.

Learning-based methods show promising robustness under long-term variations.

Insights guide future development of more reliable geometric verification techniques.

Abstract

Visual loop closure detection is an important module in visual simultaneous localization and mapping (SLAM), which associates current camera observation with previously visited places. Loop closures correct drifts in trajectory estimation to build a globally consistent map. However, a false loop closure can be fatal, so verification is required as an additional step to ensure robustness by rejecting the false positive loops. Geometric verification has been a well-acknowledged solution that leverages spatial clues provided by local feature matching to find true positives. Existing feature matching methods focus on homography and pose estimation in long-term visual localization, lacking references for geometric verification. To fill the gap, this paper proposes a unified benchmark targeting geometric verification of loop closure detection under long-term conditional variations. Furthermore, we evaluate six representative local feature matching methods (handcrafted and learning-based) under the benchmark, with in-depth analysis for limitations and future directions.