diff --git a/README.md b/README.md index 906d5d9..4ba4ade 100644 --- a/README.md +++ b/README.md @@ -22,23 +22,18 @@ To run the server, the following is needed: Without Nvidia Docker, the object detection pipeline will use the CPU to run our neural networks, and WebODM will also run without the GPU. ## Installing -1. Clone the repository: `git clone -b 360video --recursive https://github.com/RoblabWh/argus.git` +1. Clone the repository: `git clone --recursive https://github.com/RoblabWh/argus.git` 2. Run `./install.sh` from the repository directory. # take a coffee, building all images need up to 20 minutes -3. Run `./argus.sh build` to reduce running time. (requires one completion of `./install.sh`) -4. In case something goes wrong, try `git submodule update --init --recursive` +3. In case something goes wrong, try `git submodule update --init --recursive` ## Running with Installation 1. Run `argus up`: this will create and start all the containers ## Running without Installation -1. Clone the repository. Make sure to use `--recursive` or `--recurse-submodules` to also get the detection and stella vslam dense submodule: `git clone -b 360video --recursive https://github.com/RoblabWh/argus.git` +1. Clone the repository. Make sure to use `--recursive` or `--recurse-submodules` to also get the detection submodule and stella vslam dense submodule. 2. Navigate to the `argus` directory. 3. Run `./argus.sh up --build` -## Stopping the Running -1. Use `Ctrl+C` to gracefully stop argus, another `Ctrl+C` will force the containers to stop. -2. Use `./argus.sh down` to remove the containers, networkts, volumes and images that got created with `argus up` - ## Utilizing the systemd Service 1. Install ARGUS. 2. Enable the service with `systemctl enable argus`. @@ -73,4 +68,4 @@ Short Demo video on YouTube: [ARGUS - Aerial Rescue and Geospatial Utility Syste - Since ARGUS is still in development, starting multiple tasks simultaneously can lead to unexpected behavior or, in rare cases, system crashes. - Currently, ARGUS primarily supports and is tested with DJI drones (DJI M30T, as well as multiple Mavic and Mavic Enterprise models). - In order to generate fast orthophotos, the UAVs gimbal should orient the camera orthogonal towards ground (-90°). -- If WebODM has many 'WARNING Bad Request: /api/token-auth/' during startup and the redirection from ARGUS does not work, try pulling a new WebODM image with './argus.sh pull --ignore-buildable' and rebuild the server with './argus.sh up --build'. +- If WebODM has many 'WARNING Bad Request: /api/token-auth/' during startup and the redirection from ARGUS does not work, try pulling a new WebODM image with './argus.sh pull --ignore-buildable' and rebuild the server with './argus.sh up --build'. \ No newline at end of file