Algorithmic scaling of textile prints for serial multicolor printing with palette reproduction using neural networks
Вантажиться...
Дата
2025
Назва журналу
Номер ISSN
Назва тому
Видавець
Анотація
This paper proposes an algorithmic method for scaling textile prints for serial multicolor production that preserves contour geometry and palette reproducibility across different product formats. The approach integrates neural network upscaling trained to maintain boundary accuracy without inventing false textural details, joint geometric transformation of all color layers in a single deformation field, and controlled palette rescaling based on measured print parameters. This coordinated process ensures invariant contours, stable optical density, and consistent color perception during repeated runs, reducing manual correction, prepress iterations, ink usage, and batch defects. Experimental verification demonstrates stable geometry and color accuracy under realistic application conditions, supporting industrial reproducibility and resource efficiency within circular economy frameworks. Future work will focus on expanding training corpora, refining protocols for serial quality assessment, and integrating feedback from online spectrophotometric monitoring.
Опис
Ключові слова
Бібліографічний опис
Kazmirchuk Y.M., Mykytiuk M.O., Molchanova M.O., Mazurets O.V. Algorithmic scaling of textile prints for serial multicolor printing with palette reproduction using neural networks. Resource-Saving Technologies of Apparel, Textile & Food Industry. Proceedings of International Scientific and Practical Conference. November 20, 2025. Khmelnytskyi, Ukraine. Pp. 314-318.