Defects detection and characterization by using cellular neural networks

In this document, a new method to detect and to characterize surface defects in mechanical parts is reported. Cellular neural networks are used as tools for the implementation of the stereoscopic vision analysis technique. Suitable applications in microscopic defect analysis (real-time processing), in various fields (e.g. aeronautics applications) are introduced, by means of the reported examples in order to validate the cellular neural networks (CNN) approach.