2022-01-06 03:51 AM
Hi,
I am doing extrinsic calibration for VL53L5 and a RGB camera, i.e., i need to locate L5's zones in the RGB image (a zone can be representaed by a quadrilaterals in the RGB image after projection).
And my question is: How should i understand the multi-zone of L5?
(1) Do they divide the FoV equally without overlap so each zone has precisely 63/8 degree (for 8x8 zone mapping)?
(2) Or do they occupy same area in the sensor board, so the corner zone cover smaller FoV whie the center zone cover larger FoV?
2022-01-06 07:23 AM
if you consider the light goes out in a 45 degree x 45 degree pyramid, and there are 8x8 zones, you get bit more than a 5x5 degree zone. Each zone covers the same area.
The 63 degree number is the diagonal of the square, and marketing uses it because it's how one specifies TV sizes.
The sensor does the math to 'flatten' the results. So if you are looking at a wall 50 cm away, all the zones will report 50 cm - even though geometry says the photons at the edges traveled more than that.
Dot lasers have a reputation for being exact, but using a laser to flood an area and looking at the return is less precise. It's going to be close, but not spatially accurate to the mm - or to a tenth of a degree.
(In a carefully controlled lab environment, with a known target you will get more precision, but not in the real world. There are too many variables.)