PhD course: Introduction to point cloud processing for forest sciences
Are you working with remote sensing and want to learn algorithms for automated point cloud processing? This PhD course offers an introduction and hands-on experience in processing point clouds from both airborne and ground-based platforms, with an application focus in forest sciences.
Start date: 18 May 2026 00:00
End date: 30 September 2026 00:00
Language: English
Last day of registration: 11 May 2026
Organiser: Department of Forest Resource Management
Location: Online, Other location
Participants in the course will learn how to implement basic algorithms for filtering, classification, clustering, and segmentation of 3D point cloud data. Core topics include:
- Filtering and classification based on point properties and local neighborhoods
- Estimation of surface normals and surface model derivation
- Generation of high-resolution raster products from point cloud attributes
- Conditional Euclidean clustering of point clouds
- Object segmentation using classical approaches and AI-based methods, with special focus on tree crown and tree stem detection
- Processing using 3D coordinates, voxels, and 2D pixel representations
The course also covers point cloud registration workflows, visualization techniques, and efficient data structures such as Quadtrees and k-d trees, providing the computational foundation for large-scale 3D data processing.
The pace of the course is 25%.
Program
The course consists of three parts.
Part 1, May 18- June 17: Approximately 3 digital meetings
Part 2, August 17-21: A physical meeting in Flämslätt meeting facility close to our study site Remningstorp in Västergötland.
Part 3, August 24-September 30: Approximately 3 digital meetings