PTT INT TOI I ST TUE TU 
Ei EX m ui d EN REC AN 
V 7 V 
FH --—-Lt--— -———4-—4-—4 AT - - --T- 
F—+ —4 m+ mF mt mt mt mt me md md mh mL bf LL 
Ed le "ed W:elc-lb-L lle 
o Oo :5 Q 
b-+-+-+-+-+?+-4-+-+- + - + — md de + =. 
F-1—---ti-t-tt- >. -— -— -T-— ++ A A 
ni exque - — -——--4----—--d-4- - 
a! A VW —O 
Va à TA 
i models 
k bridged = 
  
Fig. 4: Height control distribution H 4 for 60% sidelap 
The corresponding accuracy law is remarkable simple again. There is only a linear dependence 
between oz mean as well as oz max and the number of models i bridged without control. 
H4: oz mean — 2.5 um i (12a) 
oz mean/oz ~ 0.25 i (12b) 
oz max — 3.2 um i (13a) 
oz max/oz ~ 0.31 i (13b) 
The formulas (12) and (13) are practicly independent of the size of the block. 
3.2.3 20% sidelap versus 60 % sidelap 
In planimetry a comparison of the formulas, related to the same control distribution shows an 
improvement at a factor 1.4 to 1.6 in accuracy when 60 % sidelap is used instead of 20 95. The poorer 
the planimetric control the more the accuracy is improved. 
In height a comparison at a common base is not possible due to the different control distributions 
in blocks with 20 % sidelap and 60 %, respectively. 
Related to approximately the same number of control points the use of a 60% sidelap instead 
of 20 9% improves oz mean and oz max the more the poorer the control is. The improvement is at the 
minimum in the order of 1.4 and increases to more than 2.0. 
3.2.4 Superwide angle photography 
Taking a look at the standard deviations in table 1, put as a foundation of this study one realizes 
that the accuracy of wideangle and superwide angle models differs practically only in the coordinates 
xo and yo of the perspective centers. But also the standard deviations ox, = oy, for wide angle and 
superwide angle correspond very well as soon as one take into account the different focal lengths. 
From this point of view practically the same accuracy after blocktriangulation can be expected for 
both image angles. In fact this expectation was confirmed by some check computations. 
3.2.5 The influence of the accuracy of control onto the accuracy after blocktriangulation 
Up to now the standard deviations ocx, ocy and ocz of the geodetic control points were assumed to 
be zero. Most of the time this is not really justified, especially not in small scale aerial triangulation. 
Therefore the following 2 versions for the accuracy of control point coordinates (ocx, ocy, ocz) were 
investigated here. 
Il 
Version 1: ocx — ocy — ox oy 
dez = OZ 
Version 2: ogex = ocy = 5 0x = 5 oy 
ocz = 2 oz