Towards High-Resolution Copy-Evident Ceramic Tiles: A Deep Learning Framework for Halftoning and Watermarking

Ceramic art is essential in interior design and decoration, making exquisite ceramic tiles imposes strict requirements for inkjet printing technology. High-resolution are often produced through printing, which the input images converted into a halftone format. However, traditional binary halftoning techniques cannot produce high-resolution ensuing process. Given that processes of high-temperature firing highly complex nonlinear system, existing methods pose intractable problems, including inconsistent textures color deviations. Based on modified U-Net model error diffusion algorithm, we propose multilevel method, capable converting color-separation images. To deter copyright infringement, further apply an ad hoc invisible watermarking method In this paper, four-stage framework: (1) A self-built dataset used to solve non-convergence overfitting problems caused by unbalanced samples non-uniqueness (2) trained applied ceramic-tile (3) An improved algorithm calibrate convert predicted continuous-tone transition printing. (4) robust watermark embedding towards proposed protection. Experimental results show our methodology effective performing color-to-halftone transformation identifying copyright.