This repository holds our Dart-based implementation of the Video program. In a nutshell, our video program is responsible for managing all USB-connected cameras and streaming them to our dashboard. In more detail, it needs to:
- open and close cameras on demand
- adjust camera settings
- monitor camera status and notify the dashboard about issues
- perform JPG compression to reduce network traffic
- be resilient against cameras being unplugged and plugged into different ports
- Integrate with our autonomy platform, which needs sole unfettered access to the RealSense camera
This repository uses Dart's FFI, which means it needs to compile C/C++ code into dynamic libraries, which need to be put in the root of this repository. So far the project only uses opencv_ffi, which has setup instructions in its repository. The necessary DLL files have already been copied over for testing on Windows devices.
To run the main program, simply run dart run. To test an individual camera, run dart run :camera <camera-name>, replacing <camera-name> with either an index, like 0 or 1, or the name of a camera device, like /dev/video0 or /dev/rear-camera.