445
trackball and in the z-direction by using a footdisk or a height drum.
As well as the hardware,the DSR-1 operating system is based on a distributed concept.
All programs are working with the menu-technique and may be user-combined in the
main menu. For the project Neustadt. the following programs were permanently applied
and defined in the main menu:
- inner orientation
- relative orientation
- aerotriangulation.
Other programs, overlapping the whole project. were
- project definition
- camera calibration input
- control point input
- BLUH (see chap.3)
- MAPS 200 (see chap.4).
To find out it's accuracy the DSR-1 was examined. The plate system is moving on spindles,
thus often causing systematic errors during measurements. Therefore a calibration was
performed with an extremely accurate 120-point-grid plate. The pointing accuracy was
calculated as being # 1.0 um and the square means of the differences of the grid being
+ 1.8 to + 2.6 um. depending on the upper and lower plate in x- and y-direction. After cal-
culating a suggestive polynom function. those results were improved to * 1.4 to £ 2.0 um
(see fig.3).
ra f i * ‘ » f + — v» + > 4 4 t & -—
+ + + + + + > + *
+ + + + f + ^ + * f } /
t » # + 9 * + + ¥ Y = } 1 * * + * t P i S +
+ + + + ^. * I * e * & + + + + Ln 4 f + > + &
* , + - - * + + 4 + - * f * + = f * + 4 > à
- - * - * — + + ? - - * + > - - + - - * > -
* 4 i P > + i i i » 4 £F + + a t + + + > * /
* > + i + + * - T + * - 2 + + - > + + f + &
* t + " t + + * * A + A * * + * + + i + * ;
s x ^ + + HS au | ft x 3 e + + + = + + t *
; i / i j
* ¥ 4 * 4 i > * ; * 4 * + + + * + > + > i
: +—— 18 MICRONS +— —-19 MICRONS
Fig.3: Results of the calibration and its improvement of the lower plate.
left: calibration results: right: results after transformation with a polynom Function
Measurements with calibrated grids at different temperatures (+ 18°C to + 26°C) were
causing no significant shift and no significant changing of scale.
The improvement of systematic errors may be calculated in an on-line program.
Those differences are small. but should not be neglected for precise measurements.
Especially by using suggestive orders and methods of measurements plenty of error cau-
sing effects can be eliminated. Those effects can be time. temperature or just the rela-
tive orientation.
