倾斜摄影模型优化和动态加载的关键技术研究

Research on Key Technologies for Optimizing and Dynamically Loading Tilt Photography Models

针对元宇宙环境中使用倾斜摄影模型构建大规模三维场景时,数据量过大和实时渲染困难的问题,提出了一种模型数据优化和动态加载的方法。方法将大量的倾斜摄影顶层数据进行重新划分,并优化合并后的几何与纹理数据,从而大幅度降低了渲染的数据压力。同时,提出针对性的瓦片动态调度算法,以及基于表面细分的实时渲染优化策略,在确保原始数据不受损失的前提下,实现了高效和稳定的倾斜摄影数据处理和渲染方法。借助这一方法,开发了一套倾斜摄影场景优化和显示系统。实验表明,上述系统具有加载速度快,运行速度快,渲染效果好的特点,可以支持数百GB以至TB级别数据的调度,可以用于各种元宇宙和数字孪生场景当中。

A method for optimizing and dynamically loading model data is proposed to address the problem of ex-cessive data volume and difficulty in real-time rendering when building large-scale 3D scenes using tilt-photography models in the metaverse environment.The algorithm redivided the large amount of tilt photography top-level data and optimized the merged geometry and texture data,thus significantly reducing the data pressure for rendering.We also proposed a targeted tile dynamic scheduling algorithm and a real-time rendering optimization strategy based on surface segmentation.A tilt photography scene optimization and display system was developed with these approaches.Experiments show that the system has fast loading speed,fast running speed and good rendering effect,and can support scheduling of hundreds of gigabytes to terabytes of data,which can be used in various metaverse and digital twin scenes.