Point-Cloud Registration of Greenhouse Data based on Infrared Imagery
Motivation and Scope
The goal of this project is to evaluate the potential of infrared (IR) imagery for point-cloud registration of plant data in greenhouses. IR has some characteristics that make it appealing for point-cloud registration: it is less light sensitive than other visual sensors, and it displays different signatures on organic matters that could be used for visual cues in detection and improved registration and monitoring.
Specific Tasks:
This project will collect a dataset using 3D IR cameras and Örebro University’s specialised automated plant growing and monitoring systems. The resulting dataset will be analysed using existing point-cloud registration algorithms. Finally, the potential for this data to be used extensively for point-cloud registration in greenhouse applications will be evaluated.
Necessary Skills:
Good programming skills, with knowledge of Python being an advantage. An understanding of point-cloud registration algorithms is also an advantage.
Benefits:
Experience with programming, computer vision, and 3D data analysis in an essential industry sector which is rapidly innovating and automating.