Abstract:

Objective Aiming at the low hidden information volume, style singularity and over-reliance on functional yarns of existing anti-counterfeiting jacquard fabrics, a fabric design and corresponding information extraction method was proposed based on unconventional dot arrangement and combination. The method enables embedding of diverse information into complex jacquard patterns and achieves anti-counterfeiting information extraction from physical fabric images, providing design ideas and case references for the development of high-value anti-counterfeiting jacquard fabrics.

Method Three unconventional dot types and their generation procedure were designed. Pairwise combinations of the three types converted the pattern and its anti-counterfeiting information image into dot patterns based on the original pattern grayscale to prepare anti-counterfeiting jacquard fabric samples. Scanned sample digital images were contrasted with fabric weave images generated with single dot shape and structural similarity and Hamming distance were used to evaluate information hiding and extraction effects respectively.

Results By calculating the structural similarity between six anti-counterfeiting jacquard fabric images and three non-anti-counterfeiting jacquard fabric images as well as the corresponding fabric weave images, the results showed that the ranking of structural similarity, from highest to lowest, was Butterfly-Heart shape combination, Butterfly-Rabbit shape combination, and then Rabbit-Heart shape combination. Additionally, the interchange of dot shapes between information and non-information areas demonstrated no significant impact on the structural similarity of the fabric sample images and their weave images. The normalized Hamming distance calculation between the extracted information images from the six anti-counterfeiting jacquard fabric sample images and the original information image revealed that the Hamming distances for the two Butterfly-Heart combination designs were 0.46 and 0.44, respectively, while the other four designs have Hamming distances between 0.09 and 0.11. Furthermore, the impact of the template cutting step size and matching threshold parameters on the information extraction effect was shown. Among the three cutting step sizes, groups 1-4 of extracted information images with a step size of 16 and 8 had similar recognition rates, which were better than those with a step size of 32. However, the recognition rates for groups 5-6 of the three step sizes were not satisfactory. In terms of matching threshold values, the best matching threshold range for the four anti-counterfeiting fabric samples using the Butterfly-Rabbit combination and Rabbit-Heart combination was 0.60-0.62, while for the two samples using the Butterfly-Heart combination, the best matching threshold values were 0.70 and 0.73, respectively.

Conclusion The experimental results were evaluated based on the similarity of the dot shapes used in anti-counterfeiting jacquard fabrics. It was found that the similarity of dot shapes is directly proportional to the information hiding effect and inversely proportional to the information extraction effect. Once the dot shape is determined, an appropriate matching threshold should be set first, followed by considering the template cutting step size. The higher the similarity of the dot shapes, the higher the template matching threshold needs to be set. Typically, the template cutting step size can be chosen as half the size of the dot shape unit. By utilizing the differences in dot shapes, this method can hide and extract complex information without affecting the gradient color pattern of the jacquard fabric. The design method of anti-counterfeiting jacquard fabric not only does not require any functional yarns, but also has a certain degree of freedom in the creative design of dot shapes and their combination, providing options for the diversified design of anti-counterfeiting jacquard fabrics.

Key words: functional fabric, anti-counterfeiting jacquard fabric design, unconventional dot structure, anti-counterfeiting information hiding, anti-counterfeiting information extraction, template matching