Printed source identification by microscopic images

The research of printed source identification is generally processed by scanned images which are limited by the scanner resolution. The accuracy of source identification is also bound by this limitation. In this study, microscopic images are used for printed source identification based on its high magnification capability for detailed texture and structure information. To explore the relationship between source printers and images obtained by the microscope, the proposed approach utilizes image processing techniques and data exploration methods to calculate many important features, i.e., Local Binary Pattern (LBP), Gray Level Co-occurrence Matrix (GLCM), Discrete Wavelet Transform (DWT), Spatial filters, Wiener filter, Gabor filter and Haralick filter. Among different set of features, LBP approach achieves the highest identification rate which is significantly superior to other methods. Consequently, the proposed technique using microscopic images achieves high classification accuracy rate which show promising applications for real world digital forensics research.