Coil defect detection: Datasets & Models
Automatically detect surface defects on steel coils using object detection, flagging issues like scratches, dents, and rust before coils move downstream.
Coil defect detection
This detection workflow analyzes camera images of steel coils and locates surface defects such as scratches, edge dents, rust patches, and coating irregularities. Built on a steel coil defect dataset, it takes still images or video frames from inspection cameras as input and returns bounding boxes with defect classes so flawed coils can be flagged before moving downstream.
Try it out
Powered by general detection model
Run on custom image
Drop an image here or click to upload
Why automate coil defect detection
Manual coil inspection in mining and metals operations is slow, inconsistent, and dependent on inspector fatigue and experience, which means defects often slip through to customers or get caught only after costly downstream processing. Automating surface inspection with computer vision provides consistent, repeatable defect detection across every coil at line speed, catching flaws early when they are cheaper to remediate. It also generates a traceable record of where and when each defect appeared, helping teams pinpoint upstream process problems.
Key benefits
Teams get objective, around-the-clock defect grading that integrates with existing inspection cameras and PLC or MES systems, reducing scrap and customer rejects. The detection approach scales from a single coil line to multiple facilities, and models like RF-DETR can be retrained as new defect types emerge, so accuracy improves over time without replacing hardware.
How to use it
- Upload an image: Drop any photo into the widget above to try coil defect detection for free, right on this page.
- Preview instant results: The model detects scratch, dent, rust spot, edge crack, coating defect 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 coil defect detection
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 coil defect detection, fine-tune it on your own labeled images, then deploy to a Serverless Cloud API endpoint.
Not sure where to start?
Describe your Coil Defect Detection project to Roboflow Agent.