NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: camera sextant?
From: Marcel Tschudin
Date: 2010 Jul 8, 01:17 +0300
From: Marcel Tschudin
Date: 2010 Jul 8, 01:17 +0300
Greg, Yes, there seem to be small systematic contributions but I think your calibration measurements are quite good. Your figure represents a "calibration function" which tends to be a quadratic, not linear, as you have drawn it. They seem to be quadratic as George has indicated it in his figures showing their possible extrapolation. In my opinion your "outliers" are "normal" points indicating only the curvature. Part of the "problem" is here that we deal with digital photos. Differences can only be 0, 1, 2 etc. pixels. A possible way to improve this would consist in taking several measurements of a position in a photo and use the mean value. If we accept the "calibration function" to be quadratic, then I would assume that most measured HEIGHTS (your data points) would be within +/-1.4 pixels, corresponding to POSITIONS being measured to +/-1 pixel. The 1 pixel difference to your previous measurements may have various reasons. I tend to attribute to your learning curve ;-) It could however be an indication to use for this sort of calibration (calibration by calculated positions) observations made at different days. Marcel