Forest Damage Detection

BEETLZ

A spatial-analysis engine that finds spruce-bark-beetle damage in aerial imagery and turns it into decisions a forester can act on.

01Problem

Problem
Canopy inspection is slow, costly and late.
Spruce bark beetles cause major financial and ecological losses, and early detection is what matters. Inspecting canopies by hand — or by eye over drone footage — doesn’t scale.
Approach
Read canopy health from the image itself.
A FastAPI service uses OpenCV to measure texture (Laplacian variance), a chlorophyll proxy (mean grey and green-band reflectance) and colour histograms to separate damaged from healthy canopy.
Build
Detections you can see and question.
A React dashboard draws bounding boxes with confidence scores over high-resolution JPG, PNG or TIFF imagery, streams the raw API payload, and is designed to plug into cloud-hosted PyTorch models.
Decision
From pixels to a cost matrix.
An intervention calculator weighs infection-lock savings against preventative revenue loss, and a cross-dataset view compares feature distributions between training data and the forest being analysed.

Results