Prediction of dot gain in flexographic color printing using machine learning
Gespeichert in:
| Titel: | Prediction of dot gain in flexographic color printing using machine learning |
|---|---|
| Autoren: | Soumen Basak, Saritha P.C, Alenrex Maity, Kanai Chandra Paul |
| Quelle: | Journal of Graphic Engineering and Design, Vol 16, Iss 4, Pp 37-49 (2025) |
| Verlagsinformationen: | University of Novi Sad, 2025. |
| Publikationsjahr: | 2025 |
| Bestand: | LCC:Mechanical drawing. Engineering graphics |
| Schlagwörter: | flexography, dot gain, machine learning, regression, neural network, Mechanical drawing. Engineering graphics, T351-385 |
| Beschreibung: | This work focuses on using machine learning algorithms in the prediction of dot gain related to flexographic process color printing. The way these advanced aspects of machine learning techniques are applied can revolutionize the various uses of printing technology. The machine learning techniques can be used to a wide range of applications since they adhere to dynamic programming methodology and computational learning theory. The machine learning algorithms can generate a trained input dataset framework, allowing them to make logical and dynamic predictions and judgments based on input data. Two grades of paper substrates with varying surface textures, two levels of anilox screen rulings, and a total of 100 steps of halftone square dot percentages with 4% intervals for each process colors are selected as the experimental process variables. An algorithm for evaluating a flexographic print output response, known as Dot Gain was generated using the Python machine learning technique. For data analysis and performance evaluation, machine learning techniques such as linear regression, decision tree, random forest regression, XG (Extreme Gradient) boost regression, SVM (Support Vector Machine) regression and neural network algorithms were used. The findings of this research work demonstrate that, out of all the machine learning algorithms used in this investigation, neural network methods had the highest accuracy. The accuracy of the neural network algorithm is 96.43, 98.32, 97.01 & 95.30 respectively in the prediction of dot gain for cyan, magenta, yellow and black. |
| Publikationsart: | article |
| Dateibeschreibung: | electronic resource |
| Sprache: | English |
| ISSN: | 2217-379X 2217-9860 |
| Relation: | https://jged.uns.ac.rs/index.php/jged/article/view/2071/1549; https://doaj.org/toc/2217-379X; https://doaj.org/toc/2217-9860 |
| DOI: | 10.24867/JGED-2025-4-037 |
| Zugangs-URL: | https://doaj.org/article/2325faabdf1342599ec56f4433a5fb53 |
| Dokumentencode: | edsdoj.2325faabdf1342599ec56f4433a5fb53 |
| Datenbank: | Directory of Open Access Journals |
Full Text Finder 
Nájsť tento článok vo Web of Science | Abstract: | This work focuses on using machine learning algorithms in the prediction of dot gain related to flexographic process color printing. The way these advanced aspects of machine learning techniques are applied can revolutionize the various uses of printing technology. The machine learning techniques can be used to a wide range of applications since they adhere to dynamic programming methodology and computational learning theory. The machine learning algorithms can generate a trained input dataset framework, allowing them to make logical and dynamic predictions and judgments based on input data. Two grades of paper substrates with varying surface textures, two levels of anilox screen rulings, and a total of 100 steps of halftone square dot percentages with 4% intervals for each process colors are selected as the experimental process variables. An algorithm for evaluating a flexographic print output response, known as Dot Gain was generated using the Python machine learning technique. For data analysis and performance evaluation, machine learning techniques such as linear regression, decision tree, random forest regression, XG (Extreme Gradient) boost regression, SVM (Support Vector Machine) regression and neural network algorithms were used. The findings of this research work demonstrate that, out of all the machine learning algorithms used in this investigation, neural network methods had the highest accuracy. The accuracy of the neural network algorithm is 96.43, 98.32, 97.01 & 95.30 respectively in the prediction of dot gain for cyan, magenta, yellow and black. |
|---|---|
| ISSN: | 2217379X 22179860 |
| DOI: | 10.24867/JGED-2025-4-037 |