Image analysis as a tool to discriminate counterfeit from true 2D printed codes

Abstract

The general context of this study is to establish recommendations for the development of digital models in the framework of counterfeiting. To achieve this goal, printed 2D codes were investigated. Visual Basic tools have been developed in order to automate tasks. The present paper allows characterizing the printing process used (conventional and waterless offset); sensitive results were also obtained regarding the kind of printed substrate (coated and uncoated paper). Histograms of area classes were plotted and they revealed that the printing process induced the raise of a new class of small dots not present on the digital file. In addition, two types of counterfeiting methods were carried out and they pointed out that the histograms of the counterfeit codes were different from the original printed code, whatever the attempt of counterfeiting. Furthermore, in these cases, small dots tend to agglomerate and form new area classes of bigger size. The method developed in this study thus allows the identification of the printing process as well as the distinction of true and counterfeit2D codes.