-9- 
(b, - zd) + (b, - za.) x! + (b, - zà,) x: 
2 
2 2 2 
1 L ' 1 1 1 ? ' ' 
i * dix; + dox: * d: i + dixi. Hann Y: * ^ 
These more restricted polynomials may be determined and applied as in sections 5.1. and 3.2. 
3.4. Parameter Restitution by modeling of flight path and attitude 
The orientation parameters may be solved for directly in & parametric least squares adjustment 
if flight path and atti_tude are modelled in & deterministic manner by polynomials of the type 
(11c) (Baker & Mikhail /7/, Konecny /48/ for scanners, and Leberl /59/ for radar). 
Another approach is to assume & periodic variation for the parameters in terms of Fourier 
series expansions (Dowideit /26, 27/, Leberl /59/ for radar; Kratky /51/, Liebig /68/ for 
scanners). The presentation here is that of /48/ : 
+ a 
approx 
2x! 
* 
approx 
2x! ; 
cos . + b, sin 
x 
tc 
&pprox 
+ e 
approx 
cos = * e 
x! 4 
m 
+ f 
o 
XY is & constant time interval, appropriately chosen for the frequency range of the data. 
(11c) or (12a) permit to set up observation equations for the image coordinates observed. 
They are represented by the collinearity equations (2) derived from (3), but for purposes 
of differentiation these are better written in the form: