Seeing the trees for the forest
This 3D model of a maple was created using a new method developed by Senior Research Fellow Pasi Raumonen. He is a member of the Inverse Problems Group in the Department of Mathematics at TUT.
Researchers at TUT have developed a new method for creating 3D models that provide a snapshot of each individual tree growing in a forest. The technology delivers an unprecedented level of detail and facilitates the sustainable and cost-effective management of forests.
Forests are a precious natural resource for Finland and growing at a rate of more than 100 million cubic metres a year. If we know exactly what kinds of trees are growing in our forests, we can make informed decisions about their use and more effectively manage our forest resources.
Researchers in the Department of Mathematics at TUT have developed 3D modelling technology that allows them to rapidly collect a massive amount of data on every single tree in a forest. The new technology promises benefits for both forest owners and the forest industry.
Accurate 3D model of each tree
The researchers employ laser scanning to capture data that is subsequently converted into a 3D model. The technique allows thousands of individual trees to be 3D-mapped from dozens of different angles using a camera-like device equipped with a rotating mirror that reflects laser beams. The resulting point cloud, which contains millions of data points of each tree, is digitally processed. The model makes it possible to identify precisely which species of trees are growing in the woods and how they can be used.
”This is the world’s first operatively efficient 3D modelling method that can be successfully applied on an industrial scale. Creating a true-to-life model of an entire forest on a similar scale is not possible anywhere else in the world. We’re capable of modelling several hectares of woodland in one day,” Professor Mikko Kaasalainen says. He leads the Inverse Problems Group in the Department of Mathematics.
”Our 3D models take data analysis to a whole new level. Forest owners, authorities and civic organizations can also use the measurement data to determine the carbon stock of forests. Our method is the only way to reliably calculate these carbon stocks,” he says.
The new technology has generated widespread interest among different forest sector stakeholders.
The research group headed by Kaasalainen is also developing a 4D model to predict the growth and yield of forests. Once completed, the model will show what happens after a certain number of trees are harvested.
Big data on forests
Professor Risto Ritala of the Department of Automation Science and Engineering participates in the Forest Big Data Project. The aim is to develop more effective methods for collecting and processing forest inventory data and the growing conditions of trees.
TUT participates in the development of a national forest infrastructure through the Forest Big Data Project. The objective is to develop effective solutions for collecting and processing forest inventory data and analysing the growing conditions of trees.
“We’re laying the foundation for a new IT platform that will provide access to accurate and up-to-date forest data that is relevant for the management of forest resources, the harvesting, transportation and sale of timber, arboriculture and forest research,” says Risto Ritala, Professor in the Department of Automation Science and Engineering at TUT.
“Satellites and airborne laser scanning systems are already collecting information about our forests. Intelligent forest harvesters could also collect data while thinning planted woods for the first time. This way we could acquire detailed information on a geographical grid at which each point is an area of 16 metres by 16 metres,” Ritala describes.
TUT’s role in the Forest Big Data Project is to develop methods for combining and updating the vast amount of data collected by harvesters and airborne laser scanners and stored in extensive regional databases maintained by the Natural Resources Institute Finland.
Forest Big Data
Launched in 2014, the two-year Forest Big Data Project is part of the Data to Intelligence (D2I) Research Programme administered by the DIGILE cluster. The project is sponsored by the Finnish Funding Agency for Technology and Innovation Tekes.
The project brings together TUT, Aalto University, the Finnish Geospatial Research Institute in the National Land Survey of Finland, the University of Helsinki, the Finnish Forest Research Institute, and the VTT Technical Research Centre of Finland. Industry partners are Metsäteho, Arbonaut, Metsähallitus, Metsä Group, Ponsse, Savcor, Stora Enso, UPM and the Finnish Forest Centre.