%0 Research article
%T Impact of point cloud matching on precision and accuracy in area-based forest inventories
%A Bollandsås, Ole Martin
%A Gobakken, Terje
%A Næsset, Erik
%A Roald, Bjørn-Eirik
%A Ørka, Hans Ole
%D 2025
%J Silva Fennica
%V 59
%N 3
%R doi:10.14214/sf.25002
%U https://silvafennica.fi/article/25002
%X <p>Reliable forest inventory methods are important for informed management. The current study compared the quality of forest attribute models based on metrics from image matching point clouds, generated using various software packages, with those based on metrics from airborne laser scanning. The field- and remotely sensed data used in the analyses were collected as part of an operational forest management inventory in Norway. Results indicate that models based on point cloud data from airborne laser scanning (ALS) consistently produced smaller root mean square error values, demonstrating superior accuracy in capturing complex forest structures compared to models using image matching point clouds. While image matching offers advantages such as lower costs and broader area coverage, this data source primarily represents canopy surfaces, which complicate its use in inventories requiring detailed canopy information. Statistical analyses revealed no significant differences in model performance among various image matching software, but all being inferior to ALS. The study emphasizes the importance of selecting the appropriate source of remotely sensed data based on specific inventory needs.</p>