Novel phenotyping techniques explored for fish breeding

Computer vision and machine learning are promising fields expected to aid in developing prediction models.

December 26, 2019

2 Min Read
WUR  seabream_c79fc0f6_490x330.jpg
Wageningen University & Research

Using images of fish before slaughter, including three-dimensional images, it may be possible to develop prediction models for traits that are currently measured after slaughter, according to an announcement from Wageningen University & Research (WUR) in the Netherlands.

Computer vision and machine learning are promising fields that are expected to aid in developing these models, WUR said, and the resulting models will then be used to design a procedure for high-throughput phenotyping.

In the future, researchers with the WUR Animal Breeding & Genomics group expect that image data will be used to automatically extract information on live fish, including morphometric measures and weight, as well as predictions of fillet yield.

WUR said current data collection systems in fish breeding programs depend on manual labor, are costly and are prone to human error.

With the aim of creating novel solutions for high-throughput phenotyping from a commercial environment, the WUR researchers and students collected data from 2,000 fish in the Mediterranean that were grown in normal production conditions in sea cages.

The researchers measured slaughter traits such as bodyweight, viscera weight and fillet weight for all fish. Most of these measurements can only be taken after slaughter and, therefore, cannot be measured on selection candidates themselves, WUR said. Currently, these traits are often measured on the siblings of the candidate fish, which limits the accuracy of selection.

In addition to the weights, the researchers also took several images of the fish before slaughter, including three-dimensional images. These images will be used to develop prediction models for the traits that are measured after slaughter, WUR said.

Machine learning

The application of machine learning algorithms to take measurements on live fish is a novel phenotyping technique that is expected to improve the selection of animals by accelerating the data collection process and removing the need to slaughter the fish to measure invasive traits, WUR said.

This research was part of the European Union's MedAID and AquaImpact projects.

Subscribe to Our Newsletters
Feedstuffs is the news source for animal agriculture

You May Also Like