This is a technical question for Richard Lemieux (or others) on the A620 distance calibration.
I notice in
http://www.cooptel.qc.ca/~rlemieu/rf2_a620_calibration.html that when plotting (1/s, x) that all the lines seem to pass through roughly (.022, 0).
If this is true, wouldn't that mean that a better way to calibrate the camera would be to fit x = a1 * ((1/s) + c), where a1 varies by focal length and c is a constant?
Also, why should the parameter for d0 ever change?
Thanks.
