When developing mobile terrestrial laser scanners (MTLS) to characterize fruit tree canopies, a need arises to validate the systems. Ground truth is difficult to establish since the developed systems usually have much more resolution than the achievable with manual measurements. Two procedures are presented to validate three different MTLS developed based on LiDAR and on Kinect v2 sensors and a RTK-GNSS receiver. The first approach consists in a) creating a scene including several geometric shapes such as prisms and cylinders, b) create a digital replica of the scene to be used as ground truth to be compared with the data obtained with the MTLS. The second approach consists in a) creating a digital model of a dried and defoliated fruit tree with high definition photogrammetric techniques, b) validate it by comparison with actual branch diameter and length measurements, and c) use it as ground truth to be compared with the data obtained with the MTLS.

The raw data resulting from the MTLS scans are processed to turn them into 3D point clouds with absolute coordinates. The point clouds are then compared with the digital mesh replica of the scene containing the geometric shapes or with the validated digital model of the dried tree. The software used for this purpose is CloudCompare, which has a function to compare a point cloud with a mesh or a point cloud with another point cloud. The result of such comparisons are distance values to the mesh or to the closest point of the other cloud for each compared point of the cloud. Those distances can be considered as errors and the number of observations can be of several hundreds or thousands, according to the scanning resolution. Descriptive statistics and statistical mean comparison techniques are subsequently applied to determine how different the obtained point clouds are from the ground truth, either the geometric shapes scene or the dried tree digital model.

The validation of the developed MTLS is crucial to subsequently use them to extract relevant information from the orchards to be used in the decision making process in the framework of Precision Fruticulture/Horticulture.