| | |
| | | # GeoAI Workbench |
| | | |
| | | ## Working conventions |
| | | |
| | | - Use Python 3.12 in project environments. Every supported workflow must retain |
| | | a CPU-compatible route. CUDA is reserved for verified model or vision |
| | | computation through a fixed environment; preparation, file I/O, geometry, |
| | | export, and the console remain CPU unless separately validated. |
| | | - The local experiment console is loopback-only on a `6xxx` port and remains |
| | | independent from drone products. Do not use browser automation for visual |
| | | acceptance; request a targeted user screenshot and run non-browser checks. |
| | | - Run the console frontend build only after Vue/TypeScript/CSS or build-time |
| | | client configuration changes. Console-owned result removal must present a |
| | | server-derived plan and confirmation, and must preserve `baseData`, external |
| | | inputs, baselines, validation assets, and unrelated runs. |
| | | |
| | | 本工作区以 [opengeos/geoai](https://github.com/opengeos/geoai) 的 `geoai-py` |
| | | 为 GeoAI 工作流基础,围绕无人机和遥感数据做实验与产品验证。这里的“GeoAI |
| | | 能力”分为三层:`geoai-py` 直接提供的功能、基于它的底层地理/AI生态组合出的 |