Fish detection: Datasets & Models
Fish detection uses computer vision to automatically locate and count fish in images or video, whether on processing lines, in aquaculture tanks, or on catch inspection tables. It identifies individual fish and can distinguish species or condition for downstream sorting.
Fish detection
This fish detection workflow analyzes images or video frames from processing lines, aquaculture tanks, or catch inspection stations and returns bounding boxes around each fish it locates. It detects individual fish and can be trained to distinguish species, separate whole fish from fillets, or flag condition and defects for downstream sorting and counting.
Try it out
Powered by general detection model
Run on custom image
Drop an image here or click to upload
Why automate fish detection
Manually counting and inspecting fish is slow, tedious, and inconsistent, especially at the throughput of a commercial processing plant or aquaculture facility. Human graders fatigue, disagree on borderline cases, and cannot easily produce an audit trail of what passed inspection. Automating fish detection gives you consistent counts, objective quality checks, and a timestamped record for every frame processed. It frees skilled workers from repetitive visual tasks and lets you catch defects or miscounts before product ships.
Key benefits
A detection model runs on your existing conveyor, overhead, or underwater cameras, so you can add automated counting and sorting without rebuilding your line. RF-DETR delivers accurate, real-time fish detection that scales from a single inspection station to an entire plant, and the model improves as you add labeled images from your own environment. The result is faster throughput, more reliable grading, and reduced manual labor across food and beverage operations.
How to use it
- Upload an image: Drop any photo into the widget above to try fish detection for free, right on this page.
- Preview instant results: The model detects fish, fillet, salmon, tuna, shrimp 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 fish 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 fish detection, fine-tune it on your own labeled images, then deploy to a Serverless Cloud API endpoint.
Not sure where to start?
Describe your Fish Detection project to Roboflow Agent.