Research on Dynamic Video of Qijiang Woodblock Prints Based on AIGC Technology
Qiang Lei, Yao Lu, Qian Li, Linzhi Li, Ye Wang, Xin Qiu · 2025
As a unique folk art form in Southwest China, Qijiang New Year pictures are facing the dual challenges of static communication limitation and insufficient technical adaptation in the digital transformation. Based on the technology of Generative Artificial Intelligence (AIGC), by constructing a multi-objective loss function that integrates the semantic alignment of text, color constraint and style feature preservation mechanism, the key problems such as color deviation and motion mutation in the dynamic process of traditional woodcut are solved. Combining optical flow-oriented interpolation algorithm and physical simulation post-processing technology, the optimization of time sequence consistency and the accurate expression of cultural symbols are realized. Through the double evaluation system of quantitative indicators and user cognitive experiments, the advantages of the algorithm in sports fluency, style fidelity and narrative consistency are verified. This technology can be applied to cultural heritage communication, educational tourism and video game industry, and its cross-domain application potential needs to be reconstructed by activating young people's immersive experience.