一种三维高斯模型的分级压缩方法以及分级渲染方法
By using hierarchical compression and rendering methods, combining spatial location and semantic features to divide Gaussian meta-levels and performing differentiated processing, the rendering latency and stuttering issues of 3D Gaussian models on mobile devices are solved, achieving high efficiency, real-time visual fidelity and compression efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUANGZHOU POWER SUPPLY BUREAU GUANGDONG POWER GRID CO LTD
- Filing Date
- 2026-04-13
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies struggle to balance visual fidelity and compression ratio when compressing and rendering 3D Gaussian models, and cannot meet the real-time interaction requirements of mobile devices, resulting in rendering delays and stuttering.
A hierarchical compression method is adopted, which calculates visual importance scores based on spatial location and semantic features, divides Gaussian units into core layer, detail layer and background layer, and performs differentiated compression strategies for each layer; during rendering, redundant data is removed by view frustum clipping and depth buffering, and differentiated lighting calculations are performed.
It achieves efficient hierarchical compression and real-time rendering on mobile devices, significantly improving visual fidelity and compression efficiency, and solving the rendering latency problem on mobile devices.
Smart Images
Figure CN122049247B_ABST