UAV Forge's ROS2 package for GN&C and Aerial Imagery Object Detection.

Running tests

Running all tests

Be in the root of the repo
./imaging_tests.sh

Running a specific test

Be in the root of the repo
python3 -m unittest tests/path-to-test/thingy_test.py

e.g. python3 -m unittest tests/imgaging/integ_tests.py

Dev Container Setup (one time setup)

Ensure you have Docker and X-11 forwarding installed on your device. (google this)
Clone the repo and edit devcontainer.json accordingly for your X-11 setup. If you're on an non-linux OS you can also just comment out the lines for mounting the X socket and uncomment the lines for the "desktop-lite" below. Then by going to localhost:6080 and entering the password vscode you should be able to access a desktop environment through your browser if necessary.
In VSCode, first run python3 setup_image.py (alternatively run echo "FROM t0mmyn/uavf_2024" > .devcontainer/Dockerfile to download from a mirrored x86 image)
Then open the command palette (cmd/ctrl+shift+p) and run rebuild and reopen in dev container
To verify your setup, run run_tests.sh

Dev Container Usage

Open VScode to the repo
Open the command palette (cmd/ctrl+shift+p) and run rebuild and reopen in dev container

Sim Usage

Refer to sim_instructions.md for instructions on starting and running the simulation.

Install instructions

Install required and local Python libraries

cd into this repo's root directory.
Run:
```
pip install -e .
```
cd into the siyi_sdk submodule and pip install -e .. If the folder is empty, do git submodule init && git submodule update

Nvidia Jetson Setup

Do this AFTER doing pip install -e . If you do that after, it'll overwrite the jetson-specific packages.

Download the torch 2.1.0 wheel from here https://forums.developer.nvidia.com/t/pytorch-for-jetson/72048 and pip install it (e.g. pip install torch-2.1..0-cp<blahblah>.whl)
Build torchvision from source.

sudo apt-get install libjpeg-dev zlib1g-dev libpython3-dev libopenblas-dev libavcodec-dev libavformat-dev libswscale-dev
git clone --branch v0.16.1 https://github.com/pytorch/vision torchvision
cd torchvision
export BUILD_VERSION=0.16.1
python3 setup.py install --user

Dev container

Open this project in vscode
Install the "Dev Containers" extension
Open the command pallete (ctrl-shift-p), then search for and execute "Dev Containers: (Re-)build and Reopen in Container"
Congratulations, you get to skip all those tedious steps to install ROS 2 manually, and your environment is isolated from the rest of your computer
To make downloading dependencies reproducible, add any important software installation steps to the Dockerfile in this repo.
To use git inside the docker container, you may have to manually log in to GitHub again if the built-in credential forwarding isn't working. I recommend using the GitHub CLI to do this.
If you want to use the simulator:
1. Follow instructions in sim_instructions.md.
2. If you want it to run it in a GUI, one way is using the remote desktop environment in the dev container. Open localhost:6080 in a web browser, then enter password vscode, then use the menu in the bottom left to open a terminal.
3. The X sockets should also be mounted and should work if you run xhost + on your machine.

I copied a lot of the config from this tutorial: https://docs.ros.org/en/foxy/How-To-Guides/Setup-ROS-2-with-VSCode-and-Docker-Container.html

Manual

(WIP). It's recommended to use the Dockerfile for development.

Running Imaging ROS Node

go into /home/ws
source install/setup.sh
colcon build --merge-install
ros2 run uavf_2024 imaging_node.py (this starts the service)
from another terminal, run ros2 service call /imaging_service uavf_2024/srv/TakePicture

Jetson Installation

pytorch: https://forums.developer.nvidia.com/t/pytorch-for-jetson/72048 as of jan 23, the jetson is on Jetpack 6

Rosbag playing example

cd /home/forge/ws/logs/bagfiles/rosbag2_2024_06_12-17_23_07 while true; do ros2 bag play rosbag2_2024_06_12-17_23_07.db3; done

uci-uav-forge/uavf_2024