Hi, I’ve been playing around with falling back to color mode in case ai object recognition fails. was trying to figure out how the awb() call works. The old vision api let you read the result back.
I was just curious as the colors generated by the ai vision configure utility in VEXCode are quite far off from being say a “yellow” for the mogo.
aivision::colordesc AIVision8__COLOR1(1, 62, 105, 50, 15, 0.42);
For reference I took a picture of a mogo using a DSLR using a white balance correction card and ran through lightroom to adjust the white balance (below). The image has no blue component to it at all - the mogo is very close to yellow - (a little more green than red).
Thanks,
Nick.