Fujitsu, Others Use AI for Evaluating Tuna's Fattiness

Tokyo, April 10 (Jiji Press)--Fujitsu Ltd. and others have developed a device that uses artificial intelligence technology to judge the fat content of frozen albacore tuna, a widely used indicator to determine the deliciousness of the fish.

Without relying on artisanal visual inspections, the automated inspection device makes it possible to quickly determine whether frozen tuna portions should be distributed labeled as high-quality, fatty "bintoro" tuna or for use to make processed products.

It is thus expected to help expand the distribution of albacore tuna eaten raw, the developers said Tuesday.

The companies will launch the device in Japan in June, targeting seafood processing firms and others. They aim to broaden the scope of automatic judgments to also cover other fish species with high distribution volumes, including yellowfin tuna and bonitos, enabling assessments of the freshness, texture and taste of the fish as well.

Given the popularity of Japanese food in global markets, the developers are also looking to take the device overseas.

Fujitsu developed the device jointly with Tokai University and food processing device makers based in the central Japan prefecture of Shizuoka.

After measuring wave patterns of fish meat by means of ultrasound analysis, the device uses AI technology to determine the fat content of the fish.

Compared with the conventional tail-cut grading process of visually inspecting the cross-section of the fish tail, the device helps reduce the time needed for the inspection by up to 80 pct and allows for complete inspections.

END