As part of the TESORO Project, financed by MIMIT, ISS developed a multispectral vision system capable of integrating the sorting function and quality control of sliced vegetables.
In particular, the system developed as part of the project evaluates the quality of each individual vegetable and any overlapping, thus providing the robot with the coordinates for picking up non-conforming ones and repositioning overlapping ones. This was achieved by integrating a 3D vision system with 2D multispectral vision systems and applying algorithms based on deep learning and machine learning.
The system was fully developed and tested at the production site of one of the partners, offering benchmark performance.

Related post