search

Team-4536 / RobotTemplate

Basic template for FRC 4536 Robots, utilizing the 2020 control system.
1 stars 0 forks source link

7 Vision Processing #2

Open iCalculated opened 5 years ago

iCalculated commented 5 years ago

Main issue at this point is getting information from the camera to the Driver Station and back.

jpwalbran commented 5 years ago

Maybe consider base-64 string encoding. Resources here: https://www.baeldung.com/java-base64-image-string

jpwalbran commented 5 years ago

Taking the image file from the Robot and sending it back to the client as a string to save memory and resources for processing.

abdulhamid6516444662 commented 5 years ago

Never mind guys i was messing around

kevintran-git commented 5 years ago

Planning a redo of the driverstation-side vision processing, now that onboard processing is working. My biggest goal is to not require OpenCV's native libraries, as they were not required at all for robot-side processing. Also, the CameraServer seems to be an automatic networktable, if it is able to communicate with Shuffleboard, so I'll just be reading up on how to implement it.

kevintran-git commented 5 years ago

However, if we plan to do offboard android/pi processing, one problem would be how would we connect to NetworkTables if the driver station is already connected to the robot? If that is not possible, how would we send information though USB?

kevintran-git commented 5 years ago

Another issue is getting the camera from the offboard computer. WPILIB automatically does it for us for the robot, but I don't know the process for getting it on an android phone for example.

jpwalbran commented 5 years ago

How about 2 Network Tables? One from the RasPi to the Robot, and one from the robot to the DriverStation?

jpwalbran commented 5 years ago

Also, from a RasPi, then we can use a USB webcam plugged into the pi