Can detection: Datasets & Models
Can detection uses computer vision to locate and identify metal beverage and food cans in images or video streams. It works on production lines, in warehouses, and at retail shelves to count, classify, and track cans automatically.
Can detection
This can detection workflow processes images or video streams and draws bounding boxes around each metal beverage or food can it locates. It detects upright and tipped cans, flags dented or damaged units, and handles dense layouts such as packed cases, pallets, and retail shelf displays.
Try it out
Powered by general detection model
Run on custom image
Drop an image here or click to upload
Why automate can detection
Manually counting and inspecting cans as they move through a filling, packaging, or palletizing line is slow, error-prone, and hard to scale across shifts. In food and beverage operations, a missed dent, a mislabeled unit, or an inaccurate case count can trigger costly recalls, waste, and downstream shipping errors. Automating can detection with a vision model gives you consistent, high-speed counting and quality checks without adding manual labor. It runs continuously, flags anomalies in real time, and produces an auditable record of what passed through the line.
Key benefits
Teams get accurate can counts, defect flagging, and inventory visibility that plug into existing line cameras and PLC or MES systems without a hardware overhaul. The RF-DETR model handles crowded, overlapping cans and varied lighting well, and it scales from a single conveyor to multiple lines and facilities. You can fine-tune it on your own can SKUs and packaging to keep accuracy high as products change.
How to use it
- Upload an image: Drop any photo into the widget above to try can detection for free, right on this page.
- Preview instant results: The model detects can, dented can, aluminum can, can lid, case of cans 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 can 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 can detection, fine-tune it on your own labeled images, then deploy to a Serverless Cloud API endpoint.
Not sure where to start?
Describe your Can Detection project to Roboflow Agent.