Hello, I have to second Bruno in noting the shortcomings of the current camera model.
And my vote is also for implementing a mathematical model that well describes real life images from real life lenses. > You are thinking of y = abs(x) = sqrt( x^^2). The first derivative > of y = x is +1 everywhere. > There are no discontinuities in any of the derivatives of non-negative > powers of x. Mathematically you do not map a one-dimensional x in R1 but a two- dimensional R2 space. For convenience one uses polar coordinates radius and phi angle but one can transform back into cartesian coordinates. Now one finds that for odd coefficients this introduces the sqrt(x^2+y^2) factor and hence some higher (partial) derivatives at point (0,0) are undefined. Does it matter in practice: the experience with my Canon A610 is Yes. I find it limits the alignment quality when doing panoramas. And I even resort of not using the a and c parameter. Adding a and c locally improves alignment but at the cost of higher misalignment in areas without control points. Recently I had to measure the height profile of a reflective surface. While the lense self-calibration from partially overlapping pano photos would have worked well in principle, and the Finetuning algorithm would have been accurate enough, I estimated the abc parametrisation to be problematic and the use of solely the b parameter not accurate enough. In the end I resorted to a non-camera method. With a proper lense model the camera would have been a very accurate angle measuring device and a sequence of only a few photos would have already recorded all the necessary data including calibration data. Regards Klaus -- You received this message because you are subscribed to the Google Groups "Hugin and other free panoramic software" group. A list of frequently asked questions is available at: http://wiki.panotools.org/Hugin_FAQ To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/hugin-ptx
