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 new AI-based digital measurement method.
Project information
Project name
Tree volume measurement by AI
Project manager
Johan Fransson
Other project members
Basam Dahy, Dag Björnberg, Shafiullah Soomro, Diana Unander
Participating organisations
Katam Technologies AB (project owner), Linnaeus University, Federal University of Viçoza, Florestal Gateados Ltda, Katam Brasil Ltda, EMBRAPII - Forest Fibers Unit
Financier
Vinnova (international collaboration for innovations with actors in Brazil, in cooperation with EMBRAPII)
Timetable
1 April 2024 – 31 March 2026
Subject
Forestry and wood technology (Department of Forestry and Wood Technology, Faculty of Technology), computer and information science (Department of Computer Science and Media Technology, Faculty of Technology)
Research groups
Forestry, Wood and Building Technologies, Linnaeus University Centre for Data Intensive Sciences and Applications (DISA), The Bridge
Knowledge Environment
Linnaeus Knowledge Environment: Green Sustainable Development.
More about the project
Background
The global demand for biomass has increased significantly and is expected to continue rising until 2050. Forests play a crucial role in the transition to renewable materials and energy, but ensuring sustainable growth requires efficient and accurate forest management. Inventories are essential for optimizing forest maintenance and profitability over time.
Today, forest inventories are conducted manually using tools such as height meters, calipers, and measuring tapes. This process is time-consuming, requires experience, and can lead to inaccurate measurements that impact economic decisions. The most sought-after information is the tree volume, traditionally estimated by measuring tree diameter and height.
New AI-Based digital measurement method for prediction of tree volume
Katam's forest inventory system, used in over 80 countries, is developing a new method to measure tree volume more accurately without the need for specific height measurements. By using GoPro cameras and AI, the system can detect the tree’s stem profile up to three meters and then predict the total tree height and volume based on the shape of the lower trunk. The project will leverage AI and automated data collection to enhance measurement efficiency.
This new method makes forest inventories faster and more accurate by eliminating manual measurement errors and time-consuming processes. AI and Computer Vision enable automated analysis of tree trunks and volumes, improving data quality for forest owners and the industry.
The method is being tested within the project in Brazilian plantation forests and Swedish forests at Attsjö super test site, near Växjö, and Svartberget, near Vindeln. The goal is to compare AI-based volume measurements with traditional methods. Additionally, the research focuses on how future forest management plans can benefit from digital decision support systems and adaptive management models to address emerging climate challenges.
This innovative method enables more sustainable and cost-effective forest management by integrating into existing processes and providing the forestry industry with better and more up-to-date information on forest conditions.
The project is part of the research in Forestry, Wood and Building Technologies and Linnaeus University Centre for Data Intensive Sciences and Applications (DISA), The Bridge, and the Linnaeus Knowledge Environment: Green Sustainable Development.