The short version is that the two different modes available determine how objects are reported when using color codes, which generally no one uses for a competition robot.
color detection is done in (a version of) YUV color space, the numbers relating to u and v specify a bounding box that we transform into a 3d lookup table for detecting colors, however, there’s really no easy way to understand exactly what the numbers mean.
Thank you for the quick reply @jpearman. I wanted to get a deeper understanding of the vision sensor because I am currently trying to find a way to distinguish between the red and blue high goals, and any other objects of that color that may mess up the vision sensors readings. My teams main concern is that if someone or something blue / red stands behinds the respective goals, our use of the vision sensor will be thrown off.