Automatic Template Images Setting Method Based on the Pix2pix Deep Learning Model for Visual Inspection System

In this paper, we propose a new method for setting parameters on image inspection system automatically, by using the Deep Learning. Generally, an image processing software built into an image inspection device has parameters that must be set manually. That is to say, the quality of parameter setting depends on a degree of operator's skill or health condition and so on, and the quality of an inspection by an automatic visual inspection system depends on the parameters set by the operator. In other words, it is necessary to set the parameters stably. Therefore, in this study, the Deep Learning model accumulates know-how to set the parameters, and this learned model sets the parameters to the visual inspection device automatically. Especially, we deal with the comparing inspection system requiring to set parameters of a template image. In the experiment, it was confirmed that the template set by our method based on the Pix2pix Deep learning model was capable of automatic inspection of the same level as the template set by the operator.