Application of Artificial Neural Network to Predict the Light fastness of Prints on Paper

Abstract

The purpose of this study is to describe the lightfastness of printed samples on the paper substrate due to long-time exposure by applying artificial neural networks. The lightfastness of prints is an important characteristic for determining their print stability. The fastness properties of prints are very important to check the print durability and image stability. It is the ability to retain the color strength of prints. It may be useful for verification of printed expiry date and authenticity or validity of the product. Nowadays, customers are very much influenced by good packaging and convinced to buy the products due to the displayed information. Packaging acts as a silent salesman and hence it is of immense importance for product manufacturers. Moreover, any kind of deterioration in package print quality will affect the product sale adversely. Little work has been done to study the fastness properties of gravure prints. In this work, paper printed in the gravure printing process has been taken as the sample as it has extensive usage in food, confectionery and medicine packaging etc. The paper samples are continuously exposed in artificial lightfastness tester BGD 865/A Bench Xenon Test Chamber (B-SUN) for assessing the light fastness of Cyan, Magenta, Yellow and Black ink on paper. The spectral curves and colorimetric values(L*, a*,b*) of prints are obtained by using ocean optics spectroradiometer (DH2000BAL) before and after exposure. An Artificial Neural Network model is proposed to predict the fading behavior of the prints. The optimal model gives excellent prediction with the minimal MSE for each color and a correlation coefficient of 0.98-0.99. As a comparison, a kinetic model is also employed. The results show that ANN has a higher prediction capability comparing to the kinetic model.