Traditional Chinese architecture emphasizes the importance of doors and windows as decorative elements, designed to allow natural light while contributing to aesthetics. This research focuses on the collection and analysis of wood-carved window grille patterns from historical architectures, utilizing a systematic approach that gathered over 2,476 images. A robust evaluation standard was established by experts to classify these patterns into categories based on complexity and regional styles. Using advanced AI techniques, a pre-trained SDM model was implemented, enabling the generation of innovative window grille designs grounded in cultural significance. Performance metrics revealed that while traditional designs excelled in detail and cultural richness, generative models offered unique symmetry and abstraction. To enhance user interaction, an AI-driven tool was developed, allowing users to create and visualize new window patterns. The integration of traditional aesthetics with modern design methods offers potential in contemporary home decor, particularly in the growing market for innovative lighting solutions.
Source link
