Introduction to point cloud processing for forest sciences
Course description
This course covers the basics of processing of high-resolution point clouds from laser scanning and digital photogrammetry. The applications in forest sciences include identification of individual trees and estimation of their properties, modelling of tree stems, and modelling of forest structure.
Course facts
P000158, Introduction to point cloud processing for forest sciences, 4.5 Hp
Print syllabus
Syllabus
Level
Third cycle
Subject
Technology
Grading Scale
The grade requirements within the course grading system are set out in specific criteria. These criteria must be available by the course start at the latest.
Course language
English
Entry Requirements
Objectives
On completion of the course, the student will be able to:
- *Use algorithms for filtering and classification of point clouds *
- Use algorithms for derivation of surface models from point clouds
- Use algorithms for segmentation of objects from point clouds
- Use algorithms for classification and interpretation of objects segmented from point clouds
Content
The course includes lectures, exercises, and student projects.
*The course introduces and teaches algorithms for automated point cloud processing. **The processing can be applied to point clouds from airborne and ground-based platforms. *An important part of the course consists of exercises where the students apply the algorithms to real data from forest sciences.
The algorithms include filtering and classification of point clouds based on properties of the points or the local surrounding, calculation of surface normals from point clouds, derivation of surface models from height data of point clouds, derivation of high-resolution raster from various properties of point clouds, conditional Euclidean clustering of point clouds, and segmentation of objects from point clouds using classical methods and AI tools, in particular tree crowns and tree stems. The algorithms are using 3D coordinates, 3D voxels, and 2D pixels.
The course also presents systems and processes to register point clouds, visualization of point clouds, data structures such as Quadtree and k-d-tree, and applications of point cloud processing in forest sciences.
Examination Formats and Requirements for Passing the Course
*Passed written exercises. Passed student project. *
Responsible Department/Equivalent
Department of Forest Resource Management