Unique log identification through artificial intelligence

The aim of this project was to develop deep learning methods for uniquely identifying individual logs through photographs. Based on the dataset collected and utilized in the project, the approach appears to work reasonably well and has potential to become a real application in the forest industry. We are currently applying for funding in order to test the method in a real operational setting. Identifying individual logs would support he legal and sustainable use of forests as well as decrease the climate impact of forestry through increasing the attractiveness and resource efficiency of lumber.