You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Sep 21, 2020. It is now read-only.
This was demonstrated at the computer vision seminar at the FRC Championships St. Louis. An audience member connected to the robot's radio, went to the address read aloud by the presenters, and started messing with the settings to hamper the ability of the GRIP pipeline that was demonstrated to find the water bottle's label.
Since the values did not update live, the presenters came to the conclusion that the lighting changed, that auto-exposure was too strong, and that the camera was faulty.
The presenters could have been properly informed about the situation if the settings were live updating, and politely asked the audience to stop changing the camera settings.
The text was updated successfully, but these errors were encountered:
It was an individual in the audience sitting in front of me.
From my perspective, it is important for sliders to live update as teams with multiple programmers may disagree on how to edit the settings to maximise the accuracy of their vision processing algorithm. Each individual may change the settings to maximise the accuracy, but only one programmer has the correct numbers on their screen (as long as they were the last to change the settings). If for some reason, the settings are reverted to default, and a programmer is called down to the pits to fix the settings, it is very likely that the incorrect settings will be applied.
From a developer standpoint, it should be fairly minimally modified to get the settings to update live given the camera image updates live.
This is a breaking change to dependencies that use the static ntcore
library. Unless the wpiutil library is also linked, linker errors will
result. This does not affect the shared ntcore library.
Sign up for freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
This was demonstrated at the computer vision seminar at the FRC Championships St. Louis. An audience member connected to the robot's radio, went to the address read aloud by the presenters, and started messing with the settings to hamper the ability of the GRIP pipeline that was demonstrated to find the water bottle's label.
Since the values did not update live, the presenters came to the conclusion that the lighting changed, that auto-exposure was too strong, and that the camera was faulty.
The presenters could have been properly informed about the situation if the settings were live updating, and politely asked the audience to stop changing the camera settings.
The text was updated successfully, but these errors were encountered: