Optical inline quality assessment of deep-drawn paperboard containers

Abstract

Presently the lack of methods for fast quality evaluation of deep-drawn paperboard containers limit the industrial applicability of the deep-drawing process. This work presents an inline method for quality evaluation of deep-drawn paperboard containers. Two images of the deep-drawn sample are taken immediately after forming within the forming machine. The images are scanned and compared with a template of an ideal wrinkle structure. Calculation of the local cross-correlation of image sections and the template allows the detection and localization of wrinkles in the wall section of the deepdrawn paperboard container. The quantity of detected wrinkles and the evenness of the wrinkle distribution on the sample can be used as a quality measure. The accuracy of the method described in this work is validated against previously applied methods for wrinkle detection. The new method outperforms the previous methods in evaluation speed by a factor of 50 without significant loss of accuracy and thus is potentially applicable for laboratory and industrial applications.