Field segmentation: Datasets & Models
Field segmentation uses instance segmentation to outline and delineate individual agricultural fields from aerial or satellite imagery. It produces pixel-level masks for each field boundary so you can measure area, count parcels, and track land use.
Field segmentation
This workflow performs agricultural field detection using instance segmentation to generate pixel-level masks for each individual field, plot, or parcel in an image. It accepts drone, aircraft, or satellite imagery as input and outputs separate masks that distinguish cultivated fields from surrounding features like roads, tree lines, water, and buildings.
Try it out
Powered by general detection model
Run on custom image
Drop an image here or click to upload
Why automate field segmentation
Manually tracing field boundaries in a GIS tool is slow, tedious, and inconsistent between analysts, and it does not scale when you need to process thousands of square kilometers or refresh boundaries every season. Automating field segmentation with computer vision turns hours of hand-digitizing into seconds of inference, giving you consistent, repeatable masks you can feed directly into acreage calculations, yield models, or subsidy verification. For agriculture teams, agronomy platforms, and ag-tech providers, this means faster onboarding of new regions and reliable land-use records without a manual annotation bottleneck. The output integrates cleanly with mapping and precision-agriculture pipelines.
Key benefits
You get precise per-field masks that support accurate area measurement, parcel counting, and change detection across imaging dates. The model works with the aerial, drone, and satellite imagery you likely already collect, so it slots into existing geospatial workflows without new hardware. Because it runs as an automated pipeline, it scales from a single farm to entire regions while keeping boundaries consistent across every image.
How to use it
- Upload an image: Drop any photo into the widget above to try field segmentation for free, right on this page.
- Preview instant results: The model detects field, crop row, tree line, farm road, water body within seconds and returns an annotated image with structured predictions, so you can immediately see how well it works. Adjust the detection classes and re-run until the results fit your scenario.
- Like what you see?: With a free Roboflow account, you can deploy this workflow with the Serverless Cloud API, customize it in your workspace, or train RF-DETR (instance segmentation) on your own images for production accuracy.
- Get started for free: No code or training data required to begin: upload, review the results, and make it part of your stack in a single click.
Recommended model for field segmentation
RF-DETR is state of the art for real-time object detection, instance segmentation, and keypoint detection, with higher accuracy and lower latency than comparable models, and it's Apache 2.0 licensed. When you're ready to build a production model for field segmentation, fine-tune it on your own labeled images, then deploy to a Serverless Cloud API endpoint.
Not sure where to start?
Describe your Field Segmentation project to Roboflow Agent.