Event
PanoAir: A Panoramic Visual-Inertial SLAM Framework and UAV Dataset
1 reports · 1 sources
The author team led by Xiangpeng Xu proposes PanoAir, a panoramic visual-inertial SLAM framework, and constructs a supporting real-world UAV panoramic dataset; the source code and dataset have been made public. According to the authors, the method completed all 17 sequences on their self-built dataset and public benchmarks without tracking failure, reducing ATE by up to 47.1% compared with the best baseline; deployment validation on an embedded platform shows that its computational efficiency is comparable to the PC implementation. The dataset covers flight conditions such as different lighting, altitudes, trajectory lengths, and motion dynamics; the method enhances feature constraints and global consistency through panoramic feature extraction and panoramic loop closure.
Report timeline
- PanoAir: A Panoramic Visual-Inertial SLAM Framework for UAVs with a Cross-Time Real-World Dataset
PanoAir proposes a panoramic visual-inertial SLAM (VI-SLAM) framework and builds a real-world panoramic dataset covering different illumination, altitudes, trajectory lengths, and motion dynamics. The method uses panoramic feature extraction and panoramic loop closure to strengthen feature constraints and global consistency. On its own dataset and public benchmarks, it completed all 17 sequences without tracking failure and reduced ATE by up to 47.1% compared with the best baseline. Deployment on an embedded platform verified computational efficiency comparable to the PC implementation. The source code and dataset are publicly available.
arXiv cs.RO ·