PhragFinder
This model is designed to identify phragmites in drone imagery, providing their correct geolocation so that environmental consultants can quickly identify the areas in need of remediation.
5 credits (with a subscription)
10 credits (without a subscription)
Version 1.0
Free Trial available!
Transform Invasive Plant Detection
Eliminate time-consuming, labor-intensive processes! Now there’s a simple, fast, and automated way to identify phragmites—with no hiking through wetlands or manually drawing polygons in satellite imagery involved.
This machine learning model uses instance segmentation to analyze drone imagery and generate geo-referenced polygons of phragmites found.
Benefits
This model will:
- Reduce the time spent looking for and treating invasive plants.
- Increase the precision of invasive plant species management.
- Help decrease the amount of funding needed for treatment of invasive plants.
- Reduce risks associated with using herbicides to control invasive plants.
Use Case
After site surveillance, run drone imagery through the model to easily identify phragmites. Model results can be fed into a sprayer drone for targeted treatment.
Required Inputs
- Georeferenced TIF files
Expected Outputs
- DXF file
- Shapefiles
PhragFinder improves the process of identifying phragmites, reducing labor time for restoration teams and allowing them to easily treat targeted areas.
Note: Current mAP (Mean Average Precision) score is 93.6%. Results will continue to be refined as the model is trained on more data.