Forest-SLAM: Autonomous navigation for closed-up Bushfire hazard mapping
UAVs technology can be used to map and monitor the bushfire hazards, so called fuel loads, in forest. The key research problem is how to fuse the forest-features with partially available GNSS information, to provide robust enough navigational information under the dense canopies. Tree-trunks are one of most prominent features to exploit, but due to the 3D/6DOP environment and vehicle dynamics, 3D tree models should be investigated. We are investigating probabilistic tree-models with branches for robust recognition. Another research problem is how to quantify the information content in the measurements and relate this information with the sensing geometry for real-time assessment.
Collaborators and Contributors:
Further Reading:
GPmap: Non-parametric Gaussian mapping
This thesis proposes new methods for robotic mapping using Bayesian nonparametric models such as Gaussian processes and Dirichlet processes. Particularly, we propose a unified framework for occupancy mapping and surface reconstruction using Gaussian processes which are called GPmaps. However, since Gaussian processes suffer from high computational complexity, it is not directly applicable to large-scale environmental mapping. Therefore, we take a divide and conquer strategy by introducing three spatial approximations, global, local and glocal (global + local) as well as a temporal approximation. The framework is being applied for closed-up mapping of ANU Solar Dishes.
Collaborators:
Further Reading:
SmartCar: Urban street mapping
This is an international collaboration project with KAIST in South Korea, aiming to build an automated 3D map fusing vision and Lidar measurements. The constructed 3D map will then be used for autonomous driving purpose.
Collaborators:
Further Reading: