top of page

Outputs

3D reconstruction of olive trees — from real-world point clouds to interactive, prunable models in Unreal Engine.

As part of the OliVR pipeline, real olive trees are captured as 3D point clouds, reconstructed into clean meshes, and brought into a real-time engine where pruning can be performed interactively — forming the basis of the project’s VR training modules.

pipeline_akis_semasi.png

For each branch segment of the tree, the following attributes were extracted: diameter, length, angle from vertical, height, topological depth, subtree size, number of children, centroid coordinates (x, y, z) and leaf count, together with a one-hot encoding of the branch class; all attributes were normalised using z-scores. 

A separate CSV file was generated for each tree, containing, for every branch segment, its coordinates, whether it should be pruned, the pruning rationale, the pruning tool to be used, the cut type and the coordinates of the cut point. 

These CSV files were produced by a rule-based decision model, and the GNN was subsequently trained on the pruning labels contained in them; the CSV is therefore the input to the GNN analysis rather than its output.

A Graph Neural Network (GNN) analysis was then performed on the resulting graph, in which spatial nearest-neighbour edges were added alongside the botanical parent–child relationships between nodes. This approach made it possible to partially recover neighbourhood-context information that is not contained in a branch's own geometry — such as "the branch is surrounded by dense branching" and "the branch is overshadowed" — whereas no such gain was observed when the botanical tree graph was used alone.

VARIETY 01 - ARBEQUINA

Arbequina Olive Tree

VARIETY 02 - DOMAT

Domat Olive Tree

outputs_eaa3b6939a.webp

VARIETY 03 - KORONEIKI

Koroneiki Olive Tree

outputs_623f177fe6.webp

INTERACTIVE - UNREAL ENGINE

Branch Pruning in Unreal Engine

outputs_47794c238b.jpg
outputs_12eac359a8.jpg

VARIETY 04 - MEMECİK

Memecik Olive Tree

Ekran Resmi 2026-07-22 20.15.26.png

VARIETY 05 - MEMECİK

memecikHastaPhoto.png

Memecik Olive Tree

memecikHasta.png

Supported By

EN_Co-fundedbytheEU_RGB_POS.png

The views and opinions expressed in this publication are solely those of the author(s) and do not necessarily reflect the views of the European Union or the granting authority. Neither the European Union nor the granting authority can be held responsible for the content of this publication or for any use that may be made of the information contained therein.

OliVR

Innovative and Digital Approaches in Fruit Cultivation:
Virtual Reality Precision Pruning for Olive Tree.

Project Coordinator

OliVR Project

© 2025-2027  OliVR project. All rights reserved

bottom of page