一种城市级实景三维轻量化建模方法
By employing adaptive optimization and GPU native format compression technology, the problem of device performance mismatch in urban-level real-scene 3D modeling was solved, achieving a reduction in data size and an improvement in performance, thus ensuring the integrity of the scene and the visual effect.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 云南省地图院
- Filing Date
- 2026-04-10
- Publication Date
- 2026-07-17
AI Technical Summary
Existing city-level real-scene 3D modeling methods fail to dynamically adjust the degree of data retention according to the performance of the target device, resulting in wasted capacity of high-performance devices and overload of low-performance devices. Furthermore, the lack of comprehensive analysis of node dependencies and semantic information affects the integrity of the scene and the visual effect.
An adaptive optimization method is adopted to construct a scene-level topology map through topology repair and hierarchical parsing. The breadth-first search algorithm is combined to traverse node dependencies, retain important landmarks, compress textures in the GPU native format, and implement perceptual hash deduplication and merging technology to generate lightweight SLPK format files.
It achieves the best balance between device performance and data scale, ensuring smooth operation of high-performance devices and fast loading speed of low-performance devices, improving the visual quality and integrity of the scene, and significantly reducing the data scale.
Smart Images
Figure CN122049277B_ABST