RGB camera-based mapping of tree stems using image matching to generate 3D models represents a low-cost, accessible approach to forest inventory that leverages standard digital cameras — including smartphone cameras and UAV-mounted RGB sensors — rather than expensive specialist instruments like LiDAR or multispectral scanners. The underlying method is photogrammetric structure-from-motion (SfM) combined with multi-view stereo (MVS) reconstruction: overlapping images captured from multiple angles around individual trees or along forest transects are processed by algorithms (such as those implemented in Agisoft Metashape, OpenMVS, or COLMAP) to identify matching feature points across images, compute camera positions, and reconstruct a dense 3D point cloud or mesh of stem surfaces. From this reconstruction, key dendrometric parameters can be extracted — diameter at breast height (DBH), stem taper, tree height, and basal area — with accuracies competitive with manual calliper measurements when image quality and overlap are sufficient. The democratisation of this technology through smartphone applications (such as single-image AI diameter estimators or multi-image SfM apps) has made it feasible for non-specialist forest workers, landholders, and citizen scientists to contribute to inventory data collection. UAV-based RGB SfM is particularly effective for mapping larger areas, producing plot-level or stand-level stem maps from above-canopy and below-canopy flight missions. In Australian forestry, RGB-based stem mapping has been trialled in both plantation and native forest contexts, with research groups at the University of Melbourne, ANU, and several state forestry agencies evaluating its accuracy against traditional cruise methods and LiDAR benchmarks. The approach is especially attractive for small-scale private forest owners and community land managers who cannot justify the cost of airborne LiDAR surveys but need accurate standing timber assessments for management planning, carbon project development, or timber sales.
| Field |
Value |
| Company |
Katam Technologies AB |
| Country |
USA |
| Type |
Patent |
| FWPA RD&E |
7.3 |
| Inventors |
Krister THAM, Linus Mårtensson, Sebastian HANER, Johannes ULÉN |
| Patent No. |
US11769296B2 |