Research
My current research concerns AI development within the forest industry, which is all part of my work as industrial PhD at Softwerk AB. In partnership with Tracy of Sweden AB we provide an algorithm for tracing of logs using the log end as a fingerprint of the log itself.
I am also currently involved in a project where we examine the possibilities of generating photorealistic images of log ends using conditional Generative Adversarial Networks (cGANs), with the ambition to create realistic training images with controlled properties.
My research groups
-
Data Intensive Software Technologies and Applications (DISTA) The research group Data Intensive Software Technologies and Applications studies data-driven approaches, such as machine learning,…
-
Linnaeus University Centre for Data Intensive Sciences and Applications The DISA research centre at Linnaeus University focuses its efforts on open questions in collection, analysis and utilization of…
My ongoing research projects
-
Doctoral project: Advanced identification methods for the forest industry through CV/AI The project intends to create opportunities for continued digitalisation in the forest industry. This concerns…
-
Project: ForestMap The overall objective of the project is to develop and evaluate a new methodology to produce forest maps across the globe, to advance the societal values of forest use. The forest…
-
Project: Tree volume measurement by AI The project is a collaboration between Sweden and Brazil with the goal to improve forest inventory efficiency in both Swedish and Brazilian forestry through a…
Publications
Article in journal (Refereed)
-
Björnberg, D., Ericsson, M., Lindeberg, J., Löwe, W., Nordqvist, J. (2024). Image generation of log ends and patches of log ends with controlled properties using generative adversarial networks. Signal, Image and Video Processing. 18. 6481-6489.
Status: Published
Conference paper (Refereed)
- Björnberg, D., Ericsson, M., Löwe, W., Nordqvist, J. (2024). Unpaired Image-to-Image Translation to Improve Log End Identification. ESANN 2024 proceedings. 673-678.