Leaf and Wood Classification Framework: 10 Global Tree Point Clouds

Ten 3D point clouds of individual trees from various global biomes, including forests and urban areas, were captured using a Riegl VZ-400 terrestrial laser scanner. This dataset was created specifically to validate a leaf and wood classification framework for terrestrial LiDAR data. The collection provides a focused benchmark for developing and testing algorithms that separate tree components.

Use Cases

• [Validating leaf/wood segmentation algorithms] based on [3D point clouds of single trees]
• [Benchmarking terrestrial LiDAR processing methods] based on [scans from diverse global biomes]
• [Training machine learning models for component classification] based on [point cloud data from a Riegl VZ-400 scanner]
• [Studying tree structure across environments] based on [data from forest and urban areas]

Strengths

• Contains 10 distinct 3D point clouds, providing a focused validation set.
• Data was collected from multiple global biomes, offering environmental diversity.
• Scans were made with a specific, high-quality terrestrial laser scanner (Riegl VZ-400).

Limitations

• The dataset is small in scale, containing only 10 point clouds.
• Key metadata such as column names, file formats, row counts, and exact update dates are unavailable.
• The specific geographic locations and time range of data collection are not detailed.

Provenance

Source

Matheus Boni Vicari, University College London

Collection Method

Scanned using a Riegl VZ-400 terrestrial laser scanner.

Geography

Global (trees scanned around the globe from different biomes, including forest and urban areas)