An Anti-Counterfeiting Method of High Security and Reliability Based on Unique Internal Fiber Pattern of Paper

A Robust Document Identification Framework through f-BP Fingerprint

“…Since in real scenarios this registration strategy is not applicable, a different acquisition framework is needed. Based on the same filter of [ 11 ], Chen et al in [ 23 ] exploited the microscopic features of wood fibers to obtain similar patterns, using expensive equipment based on double cameras. As already demonstrated in [ 24 ], the random disposition of wood fibers on paper sheets makes possible the construction of a fingerprint virtually impossible to tamper with; hence, given the limits of the previous works in terms of costs, acquisition constraints, and robustness, in [ 15 ] the authors presented a novel fingerprint extraction strategy using specific low-cost image-acquisition equipment and a simpler and faster method based on local binary patterns.…”

“…In this paper, further experiments on LBP variants are carried out, such as Local Ternary Pattern (LTP) [ 25 ], Statistical Binary Pattern (SBP) [ 26 ] and Complete modeling of Local Binary Pattern (CLBP) [ 27 ] to find the most descriptive binary pattern for fingerprint identification and tampering to achieve a more robust solution. In contrast to [ 9 , 10 , 16 , 17 , 18 , 19 , 20 , 23 ], the present work proposes a fingerprint extraction method that does not require expensive industrial equipment (e.g., laser, microscopes), but solely cheaper devices such as an RGB camera, as it is based on wood fiber translucent patterns. Additionally, the proposed approach overcomes a method based on translucent patterns such as the one by [ 11 ]; in fact, LBP descriptors used in [ 15 ] have already been proved to outperform Gabor filters employed in [ 11 ] in terms of effectiveness and efficiency.…”