Algorithm and Tool Development for Creative Generation of Graphic Design of Folk Houses and Ancient Buildings Integrating Cultural and Creative Elements

Abstract

This study presents an innovative approach called CNN-GA for graphic design for folk houses and ancient buildings, which integrates Convolutional Neural Networks (CNN) with Genetic Algorithms (GA) to foster the creation of culturally rich and aesthetically appealing graphic designs in architecture. Our research focuses on capturing the essence of folk houses and ancient buildings, deeply rooted in cultural heritage, and reimagining them through a modern computational lens. The CNN component of our model is trained on a diverse array of architectural imagery, enabling it to effectively recognize and categorize key elements such as motifs, textures, and structural forms inherent to various architectural styles. This neural network acts as an intelligent extractor of cultural and aesthetic features, providing a nuanced understanding of traditional architectural elements. The extracted features are then input into a GA, which embarks on an evolutionary process of design generation. This process iteratively combines and refines the architectural elements, fostering a creative exploration of design possibilities that maintain cultural integrity while introducing innovative interpretations. The synergy of CNN and GA in our CNN-GA framework allows for an automated yet insightful design process, yielding graphic designs that are not only architecturally sound but also resonate with the rich cultural narratives of folk houses and ancient buildings. This research holds significant potential in revolutionizing architectural graphic design, offering a novel tool for architects and designers to merge traditional aesthetics with contemporary design paradigms.