Metal surface inspection: Datasets & Models
Automated visual inspection of metal surfaces to detect and locate defects like scratches, pitting, and cracks on sheet metal, castings, and machined parts.
Metal surface inspection
This detection workflow analyzes images of metal parts to find and locate surface defects, drawing bounding boxes around issues like scratches, pitting, rolled-in scale, inclusions, and cracks. It accepts photos of steel, aluminum, castings, or machined parts from line-scan or area cameras, making it suitable for building a metal surface defect dataset and automating quality control inspection.
Try it out
Powered by general detection model
Run on custom image
Drop an image here or click to upload
Why automate metal surface inspection
Manual surface inspection in manufacturing and QC is slow, inconsistent, and hard to scale across shifts, inspectors fatigue, subtle pitting or hairline cracks get missed, and defect data is rarely logged in a structured way. A detection model applies the same criteria to every part at line speed, produces an auditable record of what was found and where, and catches defects that are easy to overlook by eye. This means fewer defective units reaching customers and earlier detection of process drift on the rolling mill or casting line.
Key benefits
Automating metal surface inspection delivers consistent defect coverage, per-part traceability, and objective pass/fail decisions that reduce scrap and rework. The workflow runs on existing inspection cameras and integrates with PLCs and MES systems, and it scales from a single station to multiple production lines without adding headcount. As new defect types appear, the underlying dataset and RF-DETR model can be retrained to keep pace.
How to use it
- Upload an image: Drop any photo into the widget above to try metal surface inspection for free, right on this page.
- Preview instant results: The model detects scratch, crack, pitting, inclusion, scale 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 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 metal surface inspection
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 metal surface inspection, fine-tune it on your own labeled images, then deploy to a Serverless Cloud API endpoint.
Not sure where to start?
Describe your Metal Surface Inspection project to Roboflow Agent.