REPORT OF COMMISSION V GV-5 
\ 
(2) Centering of the circles with respect to these axes (circle eccentricity). 
(3) Orientation of reference plane and point. 
(4) Orthogonality of the axes with respect to each other. 
(5) Vertical deviation of the telescope tube caused by gravity alone. 
(6) Dynamic deviations caused by constant and accelerated movements in 
tracking. 
(7) Deviations caused by differential expansion and contraction of compo- 
nents arising from temperature change. 
Errors arising from (1), (2), (3), (4), and (5) may be determined from star 
exposure data. The plate bubbles and vertical index constitute the horizontal 
reference plane and point. 
The accuracy of camera coordinate data is dependent on: 
(1) Magnitude and direction of image displacement owing to aberration 
characteristics of the lens. 
(2) Focal length accuracy. 
(3) Coincidence of the fiducial axes intersection with the point where the 
optical axis pierces the focal plane. 
(4) Lens system inclination. 
(5) Lens element decentering. 
(6) Film flatness. 
(7) Film distortion. 
Because of the long nominal focal length of 600 millimeters and the narrow 
cone angle of 3? by 2? errors related to (4), (6), and (7) are trivial if they fall 
within the range of similar errors found in most metrical cameras. Errors arising 
from (7) are independent of the optical mechanical system. Tests of comparable 
film exhibit maximum image displacement owing to film distortion of 11 mi- 
crons at the outer edges and 5 microns in the region of most-used imagery. 
In fact, most of the camera data errors are trivial compared to the mechani- 
cal errors caused by the length-cone-angle geometry of the camera. Perhaps the 
most significant of the camera errors are those defined by (2). The unimportance 
of camera errors is contrary to conventional photogrammetric applications 
where cameras having short focal lengths and wide cone angles are the practice. 
The difference in significance arises primarily from the fact that ballistic per- 
sonnel are interested in the coordinates of a rapidly moving object or point, 
while photogrammetrists are concerned generally in the three-dimensional 
delineation of a space containing an infinite number of static objects. 
In any case, errors defined by (2) shall be determined by star exposures in- 
sofar as (2) is significant, and (1) and (3) may be ignored for small cone angles. 
The bulk of photogrammetric effort is directed toward the delineation of an 
undulating object surface and is accomplished largely with various types of 
stereoplotting instruments. This practice has led to an opposition to an analyti- 
cal approach to the problem on the premise that it is time-consuming, compli- 
cated, and impractical. The bulk of the ballistic and missile effort is toward the 
definition of a series of space coordinates describing the path of a rapidly moving 
object, and is accomplished by analytical means with the aid of automatic 
computers. The instrumental approach is in general too slow, and inaccurate. 
Thus, independent of the photogrammetrists, mathematicians and physicists 
have developed practical procedures and methods of reducing large quantities 
of data by analytical means largely owing to the photogrammetrist's inherent 
opposition to the analytical approach. 
It is clear, then, that the method of handling photogrammetric data is 
dependent on the nature of the problem, the background of the investigators, 
and the available tools for reducing the data.