AI Fish Size Estimation Camera

Measure fish size without physical contact and manage data

The AI Fish Size Estimation Camera uses deep learning–powered image recognition to automatically measure fish size underwater or in tanks—completely contact-free. Collected data is securely managed in the cloud and can be accessed anytime from your PC or smartphone.

This system solves problems such as manual measurement errors, fish stress, insufficient sampling, and enables precise measurement that does not depend on the skill of the operator. Contactless measurement also reduces fish injury and mortality risk, improves shipping quality, and contributes to greater efficiency in aquaculture operations.

Patented technology enables precise size measurement

By utilizing image data from the proprietary stereo camera technology (patented in April 2020), highly accurate fish size measurement is achieved.

The device combines deep learning and stereo camera measurement technology to measure fish size without physical contact. Using an underwater stereo camera equipped with two lenses arranged side by side, it captures 3D images of swimming fish. Fish facing sideways are detected using deep learning object recognition technology, and the size is calculated based on the positional information obtained and stereo camera principles.

Not only body length (from the tip of the upper jaw to the base of the tail fin) and body height (from the base of the dorsal fin to the base of the pelvic fin), but also fork length and body weight can be estimated.

Analyze growth factors using data

The web browser management screen enables comprehensive viewing and management of fish size data (average, max, min, standard deviation) by group, as well as growth visualization through video storage and time-series graphs.

Real-time monitoring and data analysis from PCs or smartphones support efficient oversight of development, health, and feeding plans, enhancing aquaculture operations and management accuracy.

Fish Species Captured

Salmon, Rainbow Trout, Cherry Salmon, Yellowtail, Hamachi (Young Yellowtail),
Red Sea Bream, Striped Jack, Mackerel, Sturgeon, Carp, Grouper, Tiger Puffer, Filefish, Tilapia

Since learning models are created for each fish species, the system can measure a wide variety of species without
being limited to any particular one.