There are n matched pairs of points in 2D. How to identify outlying pairs of points according to Affine / Helmert transformation and ommit them from the transformtaion key. We do not exactly number of such outlying pairs.
I can not use Trimmed Least Squares method because there is an basic assumption that a k percentage of pairs is correct. But we do not have any information about the sample and do not know the k... In such a sample most of pairs need to be omitted (first extreme case ) or no bad pair ocurs (second extreme case).
So I tried iteratively reweightwed least squares transformation. Residuals for some points became larger, for some points smaller. Is there any algortihm using IRLS which is able to reliably detect such pairs?
Sorry for double post...