Sviluppo di un sistema di misura del colore di feltrini tessili mediante impiego di tecniche colorimetriche e reti neurali

One of the most critical phases of the production of woollen yarns for knitting and weaving is the comparison between the colour of the finished product and the colour desired by a customer or provided by a catalogue. Accordingly the companies that produce yarns perform a colorimetric control of their products by evaluating the colorimetric distance between the final product, obtained by mixing the coloured raw material, and the desired one. When this colorimetric distance exceed a value stated by the customer, the companies have to modify the combination of the raw materials in order to reduce the gap between the colour of the final product and the desired one. The objective of the present work, developed by the Dipartimento di Meccanica e Tecnologie Industriali of the University of Firenze in collaboration with the company Newmill S.p.A., is to develop an optical system able to perform an accurate colorimetric measurement and to predict the spectra of a finished yarn on the basis of the spectra of the raw materials composing the yarn itself. The developed system is based on the combination of colorimetric techniques and Artificial Neural Networks (ANNs) based algorithms.