Research interests
Photogrammetric engineering and remote sensing mapping including the development of low-cost unmanned mobile mapping systems, mapping from high resolution imagery, photogrammetric processing of image-sequences for sensor location and 3D mapping, 3D modelling using optical and lidar data and the integration of GIS and remote sensing methods and data for risk assessment and disaster management.
Research areas:
- Development and use of low-cost unpiloted mobile mapping systems (UMMS) for robotic mapping, monitoring and tracking to explore their potential and reliability in operational environments.
- UAV pose estimation (e.g., model-, RGB-D-, IMU-based navigation) in outdoor and indoor/GNSS-denied environments.
- Positioning and High-Definition mapping for Autonomous Vehicles.
- Algorithmic and computational aspects for rapid UAV-based mapping.
- Deep learning using CNN for UAV navigation and mapping.
- Alignment of 3D multi model point clouds.
- Spatial change detection.
- Rapid generation of baseline and time series data from various sensors and systems, such as air- and spaceborne imagery including unpiloted vehicle systems (UVS), satellite constellations and laser scanners.
- Development of methods and technology for spatial feature extraction, terrain modelling and spatio-temporal change detection and extraction from remotely sensed data using photogrammetry, remote sensing and GIS.
- Development of methods for the effective and efficient integration, representation, analytical visualization and delivery of spatio-temporal data and changes.
- Development of innovative geospatial applications to address emerging issues and knowledge-based decision making, such as for disaster/emergency management (geohazards monitoring, disaster prevention, disaster mitigation, and disaster response), protection of critical infrastructure, environmental monitoring, and landscape changes in the Canadian North.
Research projects
- Remotely piloted mobile mapping systems – Towards robotic mapping
- Robotic mapping, monitoring and tracking – Development of low-cost unmanned aerial mapping systems (UAMS).
- Model-based pose estimation.
- Low-cost dual IMU vision-based navigation.
- RGB-D based localization and mapping.
- Tree species classification from UAV images using of Residual Neural Networks.
- Building detection from imagery using Mask Region-based CNN.
- Spatial change detection from imagery using LeNet and Siamese CNN.
- Photogrammetric object tracking from mobile video images
- Rigid and non-rigid surface alignment
- Bluetooth based indoor positioning and navigation
- Robotic navigation and mapping using omni-directional camera
- 3D building evacuation route modelling and visualization
- Pavement surface condition evaluation using video images
- Traffic sign detection and classification from mobile lidar and image data
- Improved global web map visualization
- 3D virtual models
- Prioritization of disaster risks in a community using GIS techniques
