Inkjet printer system
AssigneeHitachi Industrial Equipment Systems Co., Ltd.
InventorsMasahiro Aono, Keisuke Okawa, Vong Vuthea
Life of this patent
Estimating the term…
Abstract
Provided is a novel inkjet printer system that can use machine learning to improve the inspection performance of a printing inspection device. The inkjet printer system includes, in the machine learning, a random test printing function unit (301) that has a function that prints a random dot arrangement that is based on various characters printed in the printing area of a printing target, and an evaluation function training function unit (302) that has a function that optimizes an evaluation function for inspecting a printed image on the basis of a captured image of the results of the printing.
Claims7
An inkjet printer system that comprises:
- an inkjet printer that performs printing onto a printing target;
- and a printing inspection device that images a printed character that is printed by the inkjet printer and performs the inspection of an imaged printed image,
- wherein the printing inspection device includes a machine learner by a neural network, and
- wherein the machine learner includes:
- a random test printer that is configured to print a dot arrangement pattern including a plurality of random dots that is based on various characters printed in the printing area of the printing target;
- and
- an evaluation function trainer that is configured to optimize the evaluation function of the neural network for inspecting the printed image on the basis of a printed image obtained by imaging the printed dot arrangement pattern,
- wherein the random dot arrangement pattern includes a normal dot arrangement pattern in which a dot is present at a normal position and a dot missing error dot arrangement pattern in which one dot is randomly selected and the selected dot is deleted from a missing portion of the normal dot arrangement pattern, and
- wherein the random test printer generates a combined dot arrangement pattern data in which the normal dot arrangement pattern and the dot missing error dot arrangement pattern are overlapped with each other such that the dot in the missing portion is emphasized and such that the neural network can be trained more efficiently.
Dependencies are inferred from the claim text.