Open Groundfish Image Library

The New England Open Groundfish Image Library provides fish images and annotations for public use.

© 2026 Integrated Monitoring, Inc. License: GNU General Public License v3.0 (GPLv3). See the NOTICE.

About the Image Library

The New England Open Groundfish Image Library was developed to support research into fish detection and species identification based on images, bounding boxes, segmentation outlines, and anatomical keypoints.

This library was developed collaboratively by Integrated Monitoring, Inc. (integratedmonitoring.net), A.I.S., Inc. (aisobservers.com), and the University of Massachusetts School for Marine Science & Technology (umassd.edu/smast).

We would like to thank NOAA Fisheries (fisheries.noaa.gov) for their support and the New England captains, crews and fisheries observers who made the collection and preparation of this dataset possible.

Funding was provided by the National Fish & Wildlife Foundation (nfwf.org) under grant #82206.

For any questions please email: info@integratedmonitoring.net

Species

Select NOAA descriptionScientific name FolderClass id EventsImages OutlinesKeypointsSize
Loading…

Preview

Only fish with
Show How much of the conveyor shows outside the outline; 0% replaces it with a checkerboard.
Keypoints

How to download images

Install the AWS CLI. No AWS account is needed: --no-sign-request makes anonymous requests. sync can be re-run to resume an interrupted download.

Everything

One species

One file type only

Browse the listing

Files are also available over plain HTTPS at <species>/<event>/<frame>.jpg (and .txt, .seg.txt, .key.txt). Reference files: classes.json, keypoints.json, length_weight.json, LICENSE, NOTICE. Per-species file lists are JSON at <species>.json, stored gzip-compressed with Content-Encoding: gzip. Browsers, Python requests and curl --compressed decompress automatically; plain curl/wget and aws s3 cp return gzip bytes, so pipe them through gunzip.

Layout and file formats

Each image is a single frame of one fish on a conveyor, with a YOLO bounding box. A subset have YOLO segmentation and keypoints. Frames are grouped by species (ITIS code) and then by capture event; several consecutive frames are usually available for each fish.

conveyor/
  classes.json                         YOLO class id -> ITIS species
  keypoints.json                       keypoint order, names and groups
  length_weight.json                   NOAA length-weight parameters by species and month
  LICENSE                              GPLv3 license text
  NOTICE                               copyright notice
  itis164744/                          species folder (ITIS TSN)
    0f3c…e2_5/                         capture event
      camera.json                                             camera calibration
      itis164744_38cm_0f3c…_20240808T015043.600000Z.jpg       image
      itis164744_38cm_0f3c…_20240808T015043.600000Z.txt       box
      itis164744_38cm_0f3c…_20240808T015043.600000Z.seg.txt   outline
      itis164744_38cm_0f3c…_20240808T015043.600000Z.key.txt   keypoints

File and folder names are identifiers used to trace each frame back to its source event; treat them as opaque. All label files use YOLO formats with coordinates normalized to 0–1 of the image width and height, one line per fish:

To train with Ultralytics YOLO, which reads <image>.txt from a labels/ folder, copy the label type you want there under the image's name, e.g. for f in */*/*.seg.txt; do cp "$f" "labels/$(basename "${f%.seg.txt}").txt"; done. For pose, add kpt_shape: [16, 3] and the flip_idx from keypoints.json to your dataset yaml.

Class ids map to species in classes.json and in the species table. Species descriptions follow the NOAA NEFSC ITIS species codes ("NK" = not known to species level).

Scroll or pinch to zoom, drag to pan, double-click to reset.

Show

Keypoints