3D imaging and machine learning improve noncontact weight estimation of frozen skipjack tuna

Determining the size and weight of fish is crucial for fisheries management and seafood processing, yet conventional methods are labor-intensive and can yield variable results among operators. Skipjack tuna caught in distant-water fisheries are often frozen on board; the frost that forms on their surfaces strongly reflects light, making precise shape measurements challenging.

To overcome this issue, researchers developed a system using a three-dimensional (3D) time-of-flight camera. This camera utilizes reflected infrared light to measure distances and capture 3D scans of frozen skipjack tuna as they move along a conveyor belt. The camera creates dense 3D point-cloud data, allowing for accurate reconstruction of the contours of frost-covered fish, even if they are frozen.

From this data, researchers extracted body width, fork length, and body height. In this study, these measurements, combined with machine-learning analysis, resulted in accurate, noncontact estimates of fish body weight that aligned more closely with measured weight classes compared to classifications made by experienced market graders.

This proof-of-concept study demonstrates the potential of 3D imaging for noncontact fish measurement and body-weight estimation. With further development toward fully automated operation, this technology could significantly reduce labor demands at fisheries and seafood-processing facilities while providing more efficient and consistent management of marine resources.