Wafer defect detection: Datasets & Models
Automated visual inspection of semiconductor wafers to locate and classify surface defects such as scratches, particles, and pattern anomalies. Uses object detection to flag failures on individual dies before packaging.
Wafer defect detection
This computer vision workflow performs wafer defect detection by processing images of semiconductor wafers from inspection microscopes, automated optical inspection stations, or line-scan cameras. It locates and classifies surface defects including scratches, particle contamination, cracks, edge chips, and pattern anomalies across the wafer and individual dies.
Try it out
Powered by general detection model
Run on custom image
Drop an image here or click to upload
Why automate wafer defect detection
Manual wafer inspection is slow, subjective, and difficult to scale across high-volume fabs where a single lot contains thousands of dies. Human inspectors fatigue, disagree on borderline cases, and cannot keep pace with throughput requirements, letting defective wafers slip downstream where rework costs multiply. Automating wafer defect detection with computer vision delivers consistent, repeatable classification at line speed and captures defect coordinates for root-cause and yield analysis. For semiconductor and pharma manufacturers held to tight quality tolerances, this means catching contamination and pattern faults earlier in the process.
Key benefits
Consistent defect classification reduces escape rates and cuts the labor tied up in tedious visual review. The workflow integrates with existing inspection cameras, AOI equipment, and microscopy hardware, so you can deploy without replacing capital equipment. RF-DETR scales across multiple inspection stations and wafer types while maintaining accuracy on small, low-contrast defects.
How to use it
- Upload an image: Drop any photo into the widget above to try wafer defect detection for free, right on this page.
- Preview instant results: The model detects scratch, particle, crack, edge chip, contamination 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 wafer 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 wafer 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 Wafer Defect Detection project to Roboflow Agent.