Bias-Eliminated PnP for Stereo Visual Odometry: Provably Consistent and Large-Scale Localization

2025-04-24

AI Summary

This work addresses the systematic bias in pose estimation within stereo visual odometry (VO) caused by triangulation uncertainty. We propose a progressively unbiased and √n-consistent weighted PnP algorithm, integrated into a decoupled spatiotemporal error modeling framework for VO. To our knowledge, this is the first PnP formulation that rigorously eliminates bias induced by noisy 3D point triangulation, with theoretical guarantees of consistency. Moreover, we introduce a novel temporal decoupling mechanism between pose estimation and 3D point reconstruction errors—enabling uncertainty-aware online tracking and triangulation as separate, modular processes. Evaluated on KITTI and Oxford RobotCar benchmarks, our method significantly reduces both relative pose error and absolute trajectory error, while maintaining robust localization performance under aggressive motion and large-scale scenes.

Technology Category

• Intelligent Robots: State Estimation
• Computer Vision: Motion & Tracking
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments

Abstract

In this paper, we first present a bias-eliminated weighted (Bias-Eli-W) perspective-n-point (PnP) estimator for stereo visual odometry (VO) with provable consistency. Specifically, leveraging statistical theory, we develop an asymptotically unbiased and $sqrt {n}$-consistent PnP estimator that accounts for varying 3D triangulation uncertainties, ensuring that the relative pose estimate converges to the ground truth as the number of features increases. Next, on the stereo VO pipeline side, we propose a framework that continuously triangulates contemporary features for tracking new frames, effectively decoupling temporal dependencies between pose and 3D point errors. We integrate the Bias-Eli-W PnP estimator into the proposed stereo VO pipeline, creating a synergistic effect that enhances the suppression of pose estimation errors. We validate the performance of our method on the KITTI and Oxford RobotCar datasets. Experimental results demonstrate that our method: 1) achieves significant improvements in both relative pose error and absolute trajectory error in large-scale environments; 2) provides reliable localization under erratic and unpredictable robot motions. The successful implementation of the Bias-Eli-W PnP in stereo VO indicates the importance of information screening in robotic estimation tasks with high-uncertainty measurements, shedding light on diverse applications where PnP is a key ingredient.

Problem

• Eliminating bias in PnP for consistent stereo visual odometry
• Decoupling pose and 3D point errors in VO pipeline
• Improving large-scale localization accuracy under erratic motions

Innovation

• Bias-eliminated PnP estimator for consistent VO
• Continuous feature triangulation for error decoupling
• Integration enhances pose error suppression