This invention relates to the field of
color printing technology, and discloses a color optimization printing method suitable for offset paper. The method includes: converting the ink droplet
diffusion process into a linear operator based on a
diffusion kernel to form a continuous image after
diffusion; defining an objective function; and alternately solving the continuous image and the objective function to output a first matrix. This invention establishes
partial differential equation constraints based on measured diffusion kernels and anisotropic parameters, ensuring that the ink volume predicted by the neural network strictly follows the actual diffusion law. Furthermore, by increasing color channels, more
layers of ink dots can be superimposed, achieving smoother color transitions. Simultaneously, it expands the color
gamut, allowing precise control of ink dot distribution in each region, ensuring clear strokes and rich colors in tiny characters, vividly reproducing image textures, shadow gradations, and other details, achieving smoother
skin tone transitions and metallic texture representation. This method is suitable for industrial-grade high-speed printing and one-pass paper printing scenarios.