Individual Tree Segmentation Using multitemporal Airborne Laser Scanning (ALS) Data

The Petawawa Research Forest in Ontario, Canada, is the location for this dataset evaluating three tree segmentation algorithms using multi-temporal Airborne Laser Scanning (ALS) data. Author Pritty Regi Arattu assessed the Dalponte2016, Li2012, and Watershed methods on ALS data from 2012, 2018, and 2022. Accuracy was measured using recall, precision, and F1-score against field-measured stem data within 14.1-meter radius plots.

Use Cases

• Benchmarking tree segmentation algorithm performance based on the described recall, precision, and F1-score metrics.
• Analyzing the temporal consistency of tree detection across multiple years based on the 2012, 2018, and 2022 ALS acquisitions.
• Studying the impact of forest structure complexity and LiDAR acquisition characteristics on segmentation accuracy as mentioned in the description.
• Comparing the over-segmentation tendency of the Watershed method against the detection patterns of Dalponte2016 and Li2012 algorithms.

Strengths

• Data spans three distinct time periods (2012, 2018, 2022), enabling longitudinal analysis.
• Algorithm performance is validated against field-measured stem data within 14.1-meter radius plots.
• Evaluation uses three standard metrics: recall, precision, and F1-score.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and file formats are unknown, which may limit suitability assessment.
• The dataset's focus is a single research forest, which may limit geographic generalizability.

Provenance

Source

Borealis Harvested Dataverse, author Pritty Regi Arattu.

Collection Method

Airborne Laser Scanning (ALS) data processed into Canopy Height Models (CHMs) for algorithmic segmentation.

Time Range

2012, 2018, 2022

Freshness

Last updated 2026-05-02 04:11:42; freshness should be verified.

Geography

Petawawa Research Forest, Ontario, Canada