112
Research
Research in the field of this commission has mainly been directed towards the development of
automatic coordinate registration equipment, the procurement of digital terrain models by
progressive sampling and the off-line production of orthophotos.
CHAPTER III. AERIAL TRIANGULATION
HT. 1.
IH.2.
Applications
The method of numerical radial triangulation has been replaced by that of independent model
triangulation for the production of topographic maps at scales 1:10.000 and 1:25.000. This
change was caused by 3 factors:
a) the accuracy of independent model triangulation is better than that of numerical radial
triangulation
b) itis nearly impossible to purchase new instruments for numerical radial triangulation
c) the limitation in the choice of terrestrial control points when using radial triangulation.
Since 1975 aerial triangulation is being carried out on the original film negatives, in which
artificial control points are made by means of stereoscopic transfer device, instead of on glass-
diapositives. Using this method has also led to a reduction in the resulting errors in orthophoto-
graphy and rectification and furthermore has eliminated the necessity for the production and
storage of glassplates.
There has also been an increase in the number of points from the national geodetic network,
such as tower spires, chimney stacks etc., which are being directly used as ground control points
in aerial triangulation.
The number of models to be triangulated has been reduced by about 209/o with the introduc-
tion of photography taken with the new standard angle camera (21/23 x 23 cm) since the
beginning of 1976.
Using this camera has also led to a reduction in the flying height from 4200 m to 3780m.
Between 1972 and 1976 a total of 55 blocks with an average of 67 models have been treated.
The resulting accuracy at a photoscale of 1:20.000 was estimated as 15 microns by calculating
the standard deviation from the block adjustment itself and 18 microns by comparing the results
of the block adjustment with special ground control points. The internal accuracy of the block
adjustment is now better than the precision of the ground control points having as a limiting
factor the identification error of these ground control points in the terrain.
Research
At the sub-department of Geodesy of the Delft University of Technology work has been carried
out on the influence that a completely filled weight-coefficient matrix of coordinates of points
measured in a model, has on the final coordinates obtained after adjustment by the "Anblock"
method.
The investigation into the precision for the determination of the projection centre in IMT by
means of the intersection of rays is still in progress.
Further developments have been made in the field of improving the process of analytical relative
orientation and in the automatic detection and elimination of gross errors in the aerial triangula-
tion phase.