In order to test this assumption, 30 points (fig. 8) were selected
to be digitized, and compared with land surveying
measurements. Twenty six points correspond to vertical
features such as walls and only four correspond in details at flat
areas. This is highly unfavourable to the orthophotos as these
points usually exhibit the largest errors. The selection was
based on the fact that the features that are commonly surveyed
and plotted represent discontinuities in surfaces rather than
‘flat’ details such as pavements and changes of tarmac. It
should be noted that although the lower photographic quality of
the orthophoto from the proposed method didn’t pose any
problem in point measurements.