A virtual scene data volume adjustment strategy based on the visual characteristics of human eyes

Abstract

Although virtual sandboxes are an essential component of platforms such as smart cities and smart parks, the efficiency of visual rendering and Web access is frequently hampered by many scenarios and models. This paper combines the polygon reduction algorithm with human visual system characteristics, proposes a real-time adjustment method for model data for virtual scenes, determines the degree of surface reduction based on the distance of the model’s center point in real-time observation, and then adjusts the number of triangular faces of the model in the scene using the collapse principle, to dynamically adjust the data amount of the model. It also uses coroutines to realize the parallel surface reduction of multiple objects and improve the rendering efficiency further. Experiments show that the optimized algorithm does not affect the visual observation effect after reducing the visual distance and storage space by 12%, indicating that it is a highly adaptive virtual scene data optimization method.