Optimization of Artificial Intelligence-based Dynamic Scene Generation and Perception Mechanism in Virtual Reality

Virtual reality, as a computer simulation system that can create virtual environments, is a fusion of multiple key technologies that are applied in different fields with important practical value and significance. This paper focuses on optimizing the artificial intelligence-based dynamic scene generation and perception mechanism in virtual reality. Firstly, a virtual dynamic scene generation method combining scene management technology and image rendering technology is proposed. In order to optimize the dynamic scene perception mechanism in virtual reality, a perception model based on spatial interaction perception and based on the analysis of user dynamic interaction is proposed. A two-layer perception management strategy is proposed. The paper finally uses the virtual reality technology proposed in this paper to create a Chinese classical garden attraction in a scenic spot. It carries out modeling accuracy assessment and comparison experiments and invites garden professionals and tourists to evaluate the overall effect of the virtual scene. The survey results show that more than 80% of the tourists are satisfied with the generation of the dynamic scene and think that it has spatial continuity and immersive experience.