The project aims to develop a machine learning system for the recognition of fish species.
In the field of intelligent manufacturing, the application context is that of technological innovation in the agri-food industry, in the fishing and aquaculture supply chain.
The project involves the prototyping of a computer vision system for the identification of a set of fish species acquired by specific IPCAMs and organized in a database distributed in Cloud Computing to allow a reinforcement learning phase of the AI model.
Particular attention is paid to the engineering and industrial automation process to ensure rapid execution times and to optimize the transmission of collected data via Cloud. The prototype will be implemented on industrial automation machines dedicated to the marketing of fish products, in one of the Italian fish markets.
The Image Analytics system foresees its use on any industrial machine where the recognition of fish species is required, allowing the automation of production processes of traceability, transformation and packaging, helping to collect further useful data downstream of the production process itself.