Design of public cultural sign based on Faster-R-CNN and its application in urban visual communication

Abstract

People are increasingly enthusiastic about pursuing spiritual life as economic and social development continues. Consequently, public cultural content has emerged as a pivotal instrument for promoting international soft power across diverse nations and regions. In today’s era of advanced artificial intelligence, cultural sign design optimization has become achievable through its deployment. This article establishes an automatic layout optimization framework, specifically tailored to meet the visual communication requirements of public cultural signage. Our framework employs Faster-R-CNN for detecting and extracting key elements of the poster, yielding an impressive average detection accuracy of 94.6%. Subsequently, we use the three-division method in design to optimize the layout, ensuring that cultural logo design conforms to visual communication principles. Our framework produced an average cultural logo satisfaction rating exceeding 70% in actual tests, providing novel insights for cultural sign design within the artificial intelligence context and significantly enhancing the efficacy of visual communication conveyed through such signage.

Keywords

• Visual Communication
• Public Cultural Signs
• Deep learning
• Object detection