Machine Learning Imaging Platform
InPhase is developing a Machine Learning Imaging Platform with funding from Innovation Norway.
This platform allows for automated acquisitions of optical and ultrasonic images to examine living fish. The system then uses machine learning and computer vision to perform classification based on discriminating parameters (e.g., fish sex or health).
A pilot application is the sorting of sedated smolt into female and male. This is achieved by detecting the ovaries using ultrasound imaging and machine learning. For this purpose, a custom-made set-up has been developed that allows for synchronous optical and ultrasonic recordings. A tailored software development tool was created to optimize the machine learning method.
If you want to know more, please contact us for further information.