If you’re just trying to view the images from a camera, this can be a great solution. The stream is embeddable in a webpage, so it integrates well with a pynetworktables2js dashboard. Includes an init script that automatically starts mjpg-streamer when the robot starts, and you can edit the settings pretty easily too.
We’re happy to announce the initial release of RobotPy for the 2016 season. This release contains most of the new changes for WPILib in 2016, with the notable exception of some new CANTalon functionality (but we hope to get that updated within the next week or so). Here’s some things that we have available for 2016:
The RobotPy project has added precompiled OpenCV 3.1.0 binaries to our 2016 opkg feed, which makes it super easy to use python + opencv on the RoboRIO. I also compiled with Java bindings, so it should be usable from C++ or Java programs if you desire also.
The pyfrc simulator now supports using real joysticks as input when you have
pygame installed for Python 3. If you have pygame installed and a supported
joystick plugged into your computer, pyfrc’s simulator will automatically
detect them and feed the joystick input to the simulator.
Because pygame can be tricky to install on some platforms, it has not been added
to the requirements for pyfrc, and pyfrc’s simulator will continue to work for
you without pygame or joysticks.
Because the communications layer uses NetworkTables, you can connect your web interface to all FRC languages (C++, Java, LabVIEW, Python).
The original working prototype for pynetworktables2js was originally created by 1418 student Leon Tan, and I’ve polished it and turned it into something that other teams can use. Team 1418 won an Innovation In Control award this weekend at the Greater DC regional, in part because of our shiny+functional HTML dashboard (look for a source code release soon!).