首个基于深度学习的全景单目视觉里程计，提升复杂场景下的定位精度与鲁棒性
360DVO: Deep Visual Odometry for Monocular 360-Degree Camera
- 设计畸变感知的球面特征提取器，自适应学习全景图像中的抗畸变特征
- 引入可微分全景束调整模块，显著提升位姿估计精度，在真实数据集上误差降低37.5%
- 构建首个真实世界全景视觉里程计评测基准，支持复杂运动与光照变化场景验证
单目全景视觉里程计（OVO）系统利用360度相机克服透视式里程计的视场限制。然而，现有方法依赖手工特征或光度目标，在剧烈运动和光照变化等挑战性场景中往往表现不佳。为此，我们提出首个基于深度学习的全景视觉里程计框架360DVO。该方法引入畸变感知球面特征提取器（DAS-Feat），从360度图像中自适应学习抗畸变特征。这些稀疏特征块被用于在新型全景可微分束调整（ODBA）模块中建立有效约束，实现精确位姿估计。为支持真实场景评估，我们还构建了一个新的真实世界OVO基准。在该基准及公开合成数据集（TartanAir V2 和 360VO）上的大量实验表明，360DVO超越当前最优基线（包括360VO和OpenVSLAM），鲁棒性提升50%，精度提高37.5%。
原文摘要 · Abstract (English)
Monocular omnidirectional visual odometry (OVO) systems leverage 360-degree cameras to overcome field-of-view limitations of perspective VO systems. However, existing methods, reliant on handcrafted features or photometric objectives, often lack robustness in challenging scenarios, such as aggressive motion and varying illumination. To address this, we present 360DVO, the first deep learning-based OVO framework. Our approach introduces a distortion-aware spherical feature extractor (DAS-Feat) that adaptively learns distortion-resistant features from 360-degree images. These sparse feature patches are then used to establish constraints for effective pose estimation within a novel omnidirectional differentiable bundle adjustment (ODBA) module. To facilitate evaluation in realistic settings, we also contribute a new real-world OVO benchmark. Extensive experiments on this benchmark and public synthetic datasets (TartanAir V2 and 360VO) demonstrate that 360DVO surpasses state-of-the-art baselines (including 360VO and OpenVSLAM), improving robustness by 50% and accuracy by 37.5%. Homepage: https://chris1004336379.github.io/360DVO-homepage
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。