Machine Learning
Tree Crown Detection
CNN and transformer models for estimating tree crown cover from PlanetScope satellite imagery.
Imagery
PlanetScope (~3m/px)
Models
CNN + Transformer
Lab
Duke Ecology
01
Overview
Tree crown cover is a key variable for carbon accounting, biodiversity monitoring, and land-use research — but field surveys cannot scale to whole regions. As an ML researcher in Duke’s Ecology Department, I build computer vision models that estimate crown cover directly from PlanetScope satellite imagery.
The work spans the full ML lifecycle: geospatial data engineering, model architecture comparison between convolutional and transformer-based approaches, and evaluation across heterogeneous terrain and canopy conditions.
02
Technical approach
- 01
Geospatial preprocessing pipeline
Raw PlanetScope scenes are tiled, co-registered with label data, normalized across acquisition dates, and filtered for cloud contamination before they ever reach a model.
- 02
CNN baselines
Convolutional encoder-decoder architectures establish strong baselines for per-pixel crown cover estimation, exploiting the local texture cues that distinguish canopy from ground.
- 03
Transformer models
Vision transformer variants capture longer-range spatial context — useful where canopy structure is ambiguous at the 3-meter resolution of PlanetScope imagery.
- 04
Robustness-focused training
Augmentation and sampling strategies are designed around generalization across terrain types, seasons, and canopy densities rather than leaderboard accuracy on a single region.
03
Architecture
01
PlanetScope Imagery
Multispectral scenes
02
Preprocessing
Tiling, normalization, QA
03
CNN / Transformer
Parallel model tracks
04
Evaluation
Cross-terrain validation
05
Crown Cover Maps
Per-pixel estimates
04
Challenges
Resolution limits
At ~3 meters per pixel, individual small crowns blur together. Models must estimate fractional cover from mixed pixels rather than detect crisp object boundaries.
Label noise
Reference data comes from higher-resolution sources with their own errors and registration offsets, so training has to tolerate imperfect supervision.
Domain shift
A model trained on one ecoregion degrades on another. Much of the engineering effort went into pipelines and augmentation that hold accuracy across varying terrain and canopy conditions.
05
Results
Working CNN and transformer crown cover models trained on large-scale PlanetScope datasets.
A reusable preprocessing and training pipeline that improved model robustness across terrain and canopy conditions.
Architecture comparison evidence informing the lab’s choice of models for region-scale mapping.
06
Media & links
More projects