Technical Innovation and Cultural Inheritance of Digital Calligraphy Art Creation

With the development of computer vision and digitization technology, automated generation of calligraphic fonts with specific styles has become a hot spot of research. Aiming at the current problems such as the strict requirements on the preliminary data collection work in the process of digitalized Chinese character art creation, this paper proposes a generative adversarial network model based on structural constraints. The structuralstyle generator, detail style generator and discriminator are used to generate, optimize and discriminate the images of calligraphic Chinese characters respectively, and finally complete the style migration to the target fonts. Analysis of the training and generation results of the model using the dataset reveals that, compared with other models, this paper's method, except for special and personalized fonts that lead to a lower recognition rate, the correct recognition rate of other fonts can be more than 70% or more, and at the same time, the digitized calligraphy art MOS value generated by the model focuses on the range of 7 to 9 points. Therefore, it can be considered that the method in this paper can simultaneously generate images of calligraphy fonts of multiple styles, and the different font styles do not affect each other, and the final generation of calligraphy fonts is also better than the existing font generation model.