Vision and artificial intelligence for leather sorting - Case Study

This project has received funding from ACCIÓ (Department of Business and Knowledge) within the framework of the competitive reinforcement grants for members of the Catalunya Clústeres 2021-2023 Programme. The aim of this collaborative project, led by Leather Cluster Barcelona with the participation of Intarex, Curtidos Badia and PICVISA, is to study the technical feasibility of a vision and artificial intelligence system for automatic leather sorting. This is a technological solution that will make companies more competitive by cutting their costs and improving their productivity. Sorting leather in the different stages of leather processing is a key factor for the quality of the manufactured product, customer satisfaction, and the operating profit of tanning companies. Today this process is performed manually, based on the expertise of workers skilled at performing the work for hours, based on mental patterns learned from accumulating great experience. They can discern between different types and combinations of defects, which allows them to sort each piece of leather into different categories. This is a very important task due to the financial impact it has on the business, since the cost of leather as a raw material is usually approximately 50% of the total manufacturing cost.

The key technology in the project is the use of vision and artificial intelligence techniques for image processing. Machine learning techniques (deep learning) will specifically be used to build computational models to identify the type of defect present in an image: parasites, scratches, cuts, etc. It is planned that in the future the project may be extended to include leather traceability.