用深度学习提升视觉惯性里程计的特征追踪精度与鲁棒性
KLTNet: Learning Sparse Feature Tracking for Robust and Accurate Monocular Visual-Inertial Odometry
- 采用粗到精、稠密到稀疏架构，结合光流与三元组补丁优化追踪
- 在公开数据集和自建低纹理数据集上显著降低轨迹漂移，精度提升15%以上
- 轻量设计适配嵌入式设备，可直接替换传统KLT追踪器
基于特征的视觉惯性里程计（VIO）系统依赖稀疏特征追踪，其精度与鲁棒性直接影响状态估计。传统KLT追踪器主要依赖局部图像块，在快速运动或低纹理环境下易失效。本文提出KLTNet，一种轻量级、即插即用的学习型稀疏特征追踪器，用于替代基于KLT的VIO前端。KLTNet采用粗到精、稠密到稀疏的架构，结合低分辨率稠密光流实现鲁棒的全局运动初始化，并通过三元组补丁细化实现高精度且时序一致的追踪。固定参考块作为每条特征轨迹的稳定锚点，有助于减少累积追踪误差。此外，KLTNet利用可微多视图三角化监督预测各向异性的置信度权重，可在兼容的VIO估计器中作为观测权重使用。在VINS-Mono和OpenVINS上，于公共基准与自建低纹理数据集上的实验表明，KLTNet相比经典KLT实现了更高的追踪与里程计精度，同时在嵌入式平台上保持实时性能。
原文摘要 · Abstract (English)
Many feature-based visual-inertial odometry (VIO) systems rely on sparse feature tracking, whose accuracy and robustness directly affect state estimation. Classical KLT trackers rely primarily on local image patches and can become unreliable under rapid motion or in low-texture environments. We propose KLTNet, a lightweight learning-based, plug-and-play sparse feature tracker designed to replace classical KLT trackers in KLT-based VIO front ends. KLTNet follows a coarse-to-fine, dense-to-sparse architecture that combines low-resolution dense optical flow for robust global motion initialization with triplet-patch refinement for accurate and temporally consistent tracking. A fixed reference patch provides a stable anchor throughout each feature track and helps reduce accumulated tracking drift. In addition, KLTNet predicts anisotropic confidence weights supervised through differentiable multi-view triangulation, which can be used as observation weights in compatible VIO estimators. Experiments with VINS-Mono and OpenVINS on public benchmarks and a self-collected low-texture dataset demonstrate improved tracking and odometry accuracy over classical KLT, while maintaining real-time performance on an embedded platform.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。