The UltracamD data set was processed by 4 network members. Some
of them used the two separate flying heights independently others
did a combined approach where low and high flying height were
considered in one adjustment approach. Since the participants
provided several versions, altogether 15 different versions were
evaluated.
The absolute accuracy was obtained from check point analysis. The
Fredrikstad test site provides 50 signalized points roughly, not all
of them visible in all flights. About 20 points were distributed as
control points. Thus about 20 were available for check point
analysis. In case the differences at some individual check point
were significantly larger than the remaining ones, the pilot centre
eliminated those point(s) from further analysis. This was done to
minimize the influence of obvious errors in image measurements on
the system accuracy analysis.
The (anonymized) results from all UltracamD participants are
given below. Please note, Ultracam low flying height corresponds to
1900m above ground (GSD 0.17m), Ultracam high corresponds to 3800m
above ground (GSD 0.34m). Besides that, the two blocks are of
different structure. Please also keep in mind the non favourable
flight conditions resulting in images of somehow reduced radiometric
quality. Details on the data itself were already given in
EuroSDR-Phase2-Data-UCDDisc.pdf ()