Discrimination of paper-based kraft tapes using Fourier transform of transmitted light images

Abstract

This study focused on two-dimensional fast Fourier transform (2D-FFT) as a new technique for the discrimination of kraft tapes, which is a kind of adhesive packing tape. The 2D power spectrum (2D-PS) obtained by applying 2D-FFT to an image enables us to obtain information about the spatial periodicity, even if the periodicity is invisible within the image. However, in the case of kraft tape, peaks in the 2D-PS are too unclear to determine its periodicity. We developed novel analytical image processes combined with 2D-FFT. 2D-FFT was applied to 50 randomly selected areas in a transmitted light image of kraft tape. The 2D-PSs were calculated from each area without applying a logarithmic transformation, accumulated, and processed by the removal of the area surrounding the center, and finally normalized for visualization. These processes enhanced the peaks and eliminated local variations. Through an intra-roll comparison, the 2D-PSs collected from a roll were similar in the location of the peaks and in their patterns at low frequency area. Using an inter-roll comparison, the 2D-PSs from 50 commercially available brand-name products were classified into 26 groups based on these peaks and patterns. All results demonstrate that this method, which is convenient, rapid, and non-destructive, could be a valuable tool for the identification of kraft tapes.