Canon Production printing develops high-end digital inkjet printer systems for the high-volume and graphic arts professional printing markets. Especially in the graphics arts market (large format, multi-pass) many different media are used: paper, PVC, canvas, textile, banner, etc.
For these multi-pass printers, the print carriage moves left-right and after each pass of the carriage, the medium makes a small step. The productivity (and quality) of the print depends on the step size. If the step size is larger or smaller than intended, this results in light or dark lines in the print and other print artifacts.
In order to reduce these print quality artifacts, printed (and scanned) markers are used to calibrate the medium and to measure and correct the media steps errors. However, correction has limitations, and it would be better to prevent the step error. This step accuracy has a large impact on the productivity and print quality of our printers.
The main goal of this PDEng project is to predict and prevent the media step size errors by using machine learning in combination with a large amount of data from a variety of data sources: printed markers, media handling sensors, temperature and humidity sensors, medium type, media settings, medium position, etc. The developed algorithm will have to control the paper steps, i.e., not just make predictions.