with the y photo axis, the higher rms error in y can be 
explained by distortions introduced by drum rotation 
anomalies. The more precise x values show that the scanner 
is more stable in the direction of the lead screw. Also 
since the discrepancies are approaching the limit of mono 
comparator measurement, the errors in manual coordinate 
measurement may account for part of the coordinate error. 
Table 1. Discrepancies between automatically generated 
coordinates and monocomparator values for photo 
#73. (All values in millimeters.) 
PANEL 
X 
y 
x-DIFF 
y-DIFF 
100 
-65.495 
81.590 
+ .014 
+ .031 
102 
-76.371 
-49.430 
-.011 
-.052 
104 
2.944 
80.873 
-.006 
+ . 022 
106 
2.263 
-59.394 
+ .011 
-.010 
110 
79.071 
24.445 
. 000 
+ .015 
114 
83.008 
-51.575 
-.007 
+ .012 
111 
81.061 
12.737 
+ .031 
+ .014 
108 
88.627 
73.111 
+ .022 
+ .012 
107 
4.786 
-65.114 
+ .007 
-.036 
105 
5.649 
80.713 
+ .033 
+ .029 
101 
-74.862 
- 1.871 
+ .011 
+ .002 
rms .017 
rms .025 
Table 2. 
Discrepancies between automatically generated 
coordinates and monocomparator values for photo 
#74. (All values in millimeters.) 
PANEL 
X 
y 
x-DIFF 
y-DIFF 
114 
- 6.844 
-54.800 
+ .016 
-.040 
111 
- 8.005 
10.092 
+ .031 
+ .031 
109 
-39.036 
29.180 
+ .019 
+ .042 
108 
0.215 
71.130 
+ .010 
+ .001 
107 
-86.871 
-68.099 
-.005 
+ .016 
105 
-84.649 
79.190 
+ .024 
+ .022 
104 
-87.402 
79.352 
+ .007 
+ .006 
106 
-89.200 
-62.266 
+ .009 
-.006 
110 
-10.000 
21.937 
.000 
+ .034 
rms .016 
rms 
.026 
A bundle adjustment was performed on the two photos using 
the automatically derived coordinates, and another adjust 
ment using monocomparator measured values was done. Table 
3 shows the results of these adjustments in terms of the 
exterior orientation parameters of the camera stations. In 
both adjustments three control points were withheld and 
used as check points. Average absolute value discrepancies 
in X, Y and Z for the monocomparator adjustment were 0.24, 
0.13 and 0.23 ft., respectively, and those for automatically 
derived photo coordinates were 0.49, 0.24 and 0.57 ft., 
respectively. These results show reasonably good agreement 
between the two techniques.