39 
s 0 is the standard error of unit weight of image coordinate measurement at the time of camera calibra 
tion and the measurement for the resection 
h is the flying altitude and c is the camera constant 
r' is the radius from the principal point to the circle upon which the points used for the calibration of 
the camera and the photographs were located 
a represents the distance from the coordinate center to the control points used for the intersection 
as follows 
x x = a x 2 = —a x 3 = —a 
Ji = 0 A 2 = —« J 3 = « 
In table 4.3:2 in the Appendix the standard errors of the elements of exterior orientation are shown 
for specific cases. 
References 4.3:1. Hallert, B.: Investigation of the Geometrical Quality of the Relative and Absolute 
Orientation Procedures and the Final Results of the Photogrammetric Procedure. 
Gimrada Res. Note No. 6, 1962. 
4.3:2. Ottoson, L.: Determination of the Accuracy of Double and Single Point Resection 
in Space. Div. of Photogrammetry Stockholm 70, 1957. 
4.3:3. Welander, E.: Some Aspects of the Accuracy of Single Point Resection in Space. Div. 
of Photogrammetry, Stockholm 70, 1958. 
f 
4.4 Tests of the Theoretical Expressions for the Accuracy of the Final Coor 
dinates of Analytical Photogrammetry 
In reference 4.4:1 a new elementary derivation of basic formula systems of analytical photogrammetry 
is presented. The final coordinates of the photogrammetric procedure are expressed as functions of the 
measured and corrected image coordinates, the elements of the interior orientation and the coordinates 
of the control points. The elementary principles used in the derivation are summarized here. Results of 
a series of practical applications are then shown. Previously derived formula systems for estimation of 
accuracy to be expected in the final coordinates are presented with values of constants taken from the 
results of the investigations of the fundamental operations mentioned earlier. The theoretical accuracy 
is compared with the practical results and the inevitable differences are statistically tested for signifi 
cance. From fig. 4.4:1 the following relations can be directly derived: 
H c i) _ ( c i) ( c 2 ) b 
(* ) ^ (c 2 ) ^ ^ (*") ( Cl ) 
(4.4:1) 
r _ (*') * .. ( C 2 ) (*') h 
( c i) (A) (c 2 ) — (*") (c x ) 
(4.4:2) 
r _ (/) * (Cs) (/) / 
(Cl) (x') (c 2 ) — (x") (c x ) 
(4.4:3) 
r _ (y") z ( c i) (y") 1 
(c 2 ) (x') (c 2 ) — (x") (Cj) 
(4.4:4)