Blueprint Segmentation with Mask R-CNN
A computer vision pipeline that uses a Mask R-CNN deep learning model to read architectural blueprints like a floor plan.
The Challenge
Indoor mapping needed a computer vision pipeline that could automatically identify walls, doors, points of interest, and text directly from raw blueprint images, without manually digitizing every building by hand.
What We Built
We annotated blueprint imagery in Roboflow and trained a Mask R-CNN instance segmentation model via Detectron2 to detect walls, doors, and points of interest directly from blueprint images, then converted blueprint pixel coordinates into real-world building coordinates using an affine transformation pipeline.
- Custom-annotated blueprint dataset built in Roboflow for training and validation
- Mask R-CNN instance segmentation model trained via Detectron2 to detect walls, doors, and points of interest
- Text recognition to extract room labels and identifiers directly from blueprint images
- Affine transformation pipeline to convert pixel coordinates into real-world building coordinates
- Structured output feeding directly into the Mapify routing engine
Where It Stands
Shipped in one month as the computer-vision backbone powering the Mapify indoor navigation apps.
Tell us about your project and we'll follow up with a scoped quote.
Get a QuoteLive Preview
This is a backend computer-vision pipeline with no standalone interface, it powers the Mapify apps rather than running as its own product.
Keep Exploring
Related reading and services
Annotating a Custom Dataset for Blueprint Segmentation in Roboflow
The dataset and training work behind our Mask R-CNN blueprint segmentation model: annotating architectural blueprints in Roboflow and training via Detectron2.
Read more →ServiceData Science & Analytics
Turning your data into models, dashboards, and decisions.
Read more →More Work