Reproduce “GPS Denied IBVS-Based Navigation and Collision Avoidance of UAV Using a Low-Cost RGB Camera” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Xiaoyu Wang, Yan Rui Tan, William Leong, Sunan Huang, Rodney Teo, Cheng Xiang
Researchers at the National University of Singapore developed a system for Unmanned Aerial Vehicle (UAV) navigation and collision avoidance in GPS-denied environments, utilizing only a low-cost RGB camera. The framework integrates Image-Based Visual Servoing (IBVS) with AprilTags for multi-target navigation and AI-driven monocular depth estimation (MiDaS) for real-time obstacle avoidance, performing all computations onboard a Jetson Orin platform.
Key insights
- Robust navigation and collision avoidance in GPS-denied settings can be achieved using the synergistic combination of relative visual features (AprilTags) and learned relative depth (MiDaS) from a single low-cost RGB camera.
- Real-time AI inference for monocular depth estimation is practical on embedded platforms like Jetson Orin, enabling full onboard autonomy without external computing resources.
- A reactive, IBVS-based control approach, coupled with thresholded depth analysis, effectively manages multi-target navigation and obstacle avoidance without requiring explicit 3D path planning or precise global localization.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.