Clarification of root architecture is important for understanding tree root functions. 3D laser scanning reconstructs the 3D structure of root system with high accuracy, but the devices are large and expensive. The Structure-from-Motion with multi-view stereo-photogrammetry (SfM-MVS) method reconstructs the 3D structure of an object from multiple images. This method is suitable for field research because it acquires the 3D data with a commercially available digital camera. Our aim in this study was to test the applicability of the SfM-MVS method for morphological measurement of tree root systems.
The 3D root system model was created using a dummy and three black pine (Pinus thunbergii) root systems. Image data were collected from the root systems using a camera and reconstructed using the free software, VisualSfM and MeshLab. We evaluated the accuracy of morphological data estimated from the 3D models by comparing them with real data obtained from manual measurements of the root systems.
The 3D reconstructions of the dummy and black pine root systems were successful. The root mean squared error values for the morphological features in the pine root systems ranged from 4 to 24%, which was sufficiently accurate for a successful 3D model. However, the RMSE values were larger than that of dummy root, ranging from 2 to 9%.
The SfM-MVS method is a new tool, based on non-specialised equipment and free software, to obtain the 3D structure of tree root systems. The morphological features of it can be successfully measured from the reconstructed models.