Because the unknowns Kj, Ko, and Kz are linear in the observational 
equations (76), their approximations in each iteration may be taken equal to 
« 
zero; consequently the coefficients of the remaining unknown orientation 
parameters remain the same as given by formula (hi). 
The incorporation of the distortion determination requires merely the 
addition of the partial differentials with respect to the unknowns Ki, K, and 
Kz to the system of observational equations. 
They are, for the x-equation: 
for  : Ky : M = a CE = a . (49 - x?) 
1 x q x 
4 em — JH O Oo 
K, :N, -d e "8 « (4. 7 X) 
K.:0 d. D (7 - x9) 
5 X q X p 
and for.the y-equation: (78) 
2 en 2 o O 
for K. : MM =a, —=4 (££ - 
i y a ( y Ip) 
| cn lh. 7,0 o 
Ks : N =4 — -d (Z4 - 
BI q Gy 7 Yo) 
6 en 6 , ,0 O 
K 0. = d — zd 2 - 
5 y q y Yo) 
where 
Ball - PR à 102 - SP CC 4 SU (for the meaning of C. and C 
T ^p y y x 
p y- x xy 
see formulas (hO) and (41)) 
Consequently we obtain: 
M 0 .€ —M o0 .C 
cvy Y zy "x 
2 2 
N =C. 6 -N =o. 6, (79) 
Q 2C .c  »Q c 
x Xy y y Xy X 
52 
ac 
wk 
ie 
tL