Overseer is designed for quickly deploying computer vision on a Raspberry Pi for the FIRST Robotics Competition.
Create a file called gradle.properties
in the root directory of this project
Set the values:
vision_user=<username>
vision_password=<password>
vision_host=<ip address of the host>
In the build.gradle, increment version
whenever a major change is made.
The GripPileline.java
is generated by GRIP and is only in this repository as an example.
It should be replaced with the intended GripPipeline.
-
Run
./gradlew build
to create the jar -
Run
./gradlew deploy
to send the jar and the launch script to the raspberry pi
-
This project works by deploying the built jar to the raspberry pi over an SFTP connection and restarting a systemd unit on the pi (
vision.service
). -
All important files are upload to
/srv/vision/
-
The systemd unit file launches the
run_vision.sh
file found in this project to launch the jar file that was uploaded -
In
/srv/vision/natives
there are some integral raspbian compiled shared objects necessary for the WPILib cameraserver, WPILib networktables and opencv to run in java. Currently these are manually put on the Pi and can be found at these locations:In the future I intend to automatically put these on to a pi on deployment