Hatchery marked otolith images and classification models

Date 2021-10-13
Author(s) Doherty Susan2, Kemp ChandlerORCID1
Affiliation(s) 1 : Kempy Energetics
2 : Otolith Marking and Reading Research
DOI 10.17882/84047
Publisher SEANOE
Keyword(s) Otolith, hatchery, otolith marking, salmon
Abstract

These data contain 250 images of hatchery marked and unmarked otolith images. The images include otoliths with one of four distinct marks. There are 50 images that contain each of the four marks, and 50 images of unmarked otoliths. The images were used to train and test neural networks for use in identifying the marks. The networks were trained on the first thirty images in each class. The remaining twenty images in each class can be used for testing. Two trained networks are included: "binarynet" distinguishes marked and unmarked images, and "classnet" classifies marked images. Two versions of "binarynet" are available: one trained on the same database as "classnet" and a second iteration stored in a "retrained" directory that was finetuned using adversarial samples selected from the training images by "classnet." Finally, a set of software utilities written in python are included that show how the networks were trained and process the images for classification by the networks.

Licence CC-BY
Data
File Size Format Processing Access
Zip archive containing images of otoliths with hatchery marks. The specific mark is indicated by the subdirectory name. 50 examples of each of four distinct marks and unmarked otoliths are included. 1 GB Zip archive of JPG files Raw data Open access
Zip archive containing trained neural networks that can be used to classify otolith images. 94 MB Zip archive of h5 and pickled python objects. Processed data Open access
Zip archive containing python utilities for processing otolith images and using the associated classification networks 535 KB Zip archive of python modules Open access
Top of the page