18 
of course an optical projection/transformation similar to 
the original orthophoto printing step. 
OFF-LINE ORTHOPHOTO PRINTER TESTS 
Two tests were conducted in the evaluation of the 
Off-Line Orthophoto Printer (OOP). The first test used an 
■thophoto produced by Bendix Research Labs from a stereo 
ir of KC-1 photography. The scucond was conducted with an 
: thophoto produced at RADC from RC8 imagery sent by the 
; fcional Research Council of Canada of their Sudbury 
; lOtogrammetric test range. 
1. Preliminary Acceptance Test - California Test Model, 
hi a Bendix test model was taken over California at a scale of 
1/200,000 by the X-15. The KC-1 camera used has a nominal 
focal length of 6 inches and 9x9 frame format. Profiles were 
compiled from the model for which only a relative orientation 
w.is established. The orthophoto was produced using a nominal 
print width of 2mm at a scale of about 1/100,000. (Figure 7). 
The stereo pair was then sent to RADC along with the orien 
tation data (shutdown tape) where the model was re-established 
on the AS-11B-1 analytical plotter. Absolute control for this 
test was not available so relative model control was derived 
and used instead. The XY coordinates of the model surface 
lying directly under 37 grid intersections were measured and ! 
recorded. The corresponding grid intersections on the ortho 
photo were then measured using one stage of the AS-11B-1 as 
a monocomparator. A least squares adjustment was used to 
translate, rotate and scale the coordinates of the orthophoto 
grid intersections to those measured in the stereo model. The 
residuals resulted in a standard deviation of 14 microns in 
X and 12 microns in Y at the scale of the input photography. 
This is equivalent to a CMAS of 37 microns (Circular Map 
Accuracy Standard, i.e., 90% probability that a point lies 
within a radius of 37u from its "true position*) . These results 
were considered very encouraging as the contract specifications 
with Bendix had called for a standard deviation of 50 microns 
in position X and Y, referred to the input. 
2. Final Acceptance Tests-Sudbury Model. 
The final acceptance test was more comprehensive in that ground 
control was used in place of model control. The Sudbury stereo ; 
pair was graciously provided by Dr. T. Blachut and his associates 
at NRC Canada. Sudbury was an excellent test model because of 
the abundant control available. The camera was an RC8 with a 
nominal focal length of 6 inches. The model orientation, profiling 
and evaluation was performed by Mr. Don Hall at RADC on the AS-11B-1