Non-destructive discrimination of white papers by the most frequent brightness value of images

Abstract

 Non-destructive discrimination of white paper sheets of 9 different notepads was tried by analyzing the color of the images. All of the paper sheets of these notepads were scanned to obtain reflection, transmission, and fluorescence images using the scanner with multiple light sources. The scanning condition was 300 ppi and 24 bit RGB color, namely 256 gradations. We examined the brightness values of the R, G, and B channels in the width of 640 pixels and the height of 1020 pixels, a card size, from each of the obtained three images.

 Since the brightness values were uneven within one paper, the most frequent values of those were used for comparison of the notepads. The most frequent values were nine kinds for one paper because each of the obtained three images had three channels. The most frequent values had values with a range in one notepad. Therefore, those ranges were nine kinds for one notepad.

 As a result, comparing nine ranges of the most frequent values of one notepad with other notepads, 9 different notepads could be discriminated from each other by some of nine ranges. Thus, this study indicates that the image analysis using the most frequent values of the brightness values is useful for non-destructive discrimination of white paper sheets in forensic science.