Of these four major elements of the system the first two are produced by GSI 
and the last two are obtained by GSI from vendors and integrated into the 
system. 
2.0 THE GSI CLOSE-RANGE CAMERA CRC-1 
2.1 INTRODUCTION 
The CRC-1 Camera (Figure 2) has been specifically designed by the 
writer to conform to his concept of the ideal camera for meeting the most 
exacting demands of close-range photogrammetry. As will presently be made 
clear, the design of both camera and software were executed in parallel, 
with each reenforcing the capabilities of the other. This is the first 
time that such a synergistic approach to hardware/software design has been 
exercised in close-range photogrammetry; the benefits are substantial. 
General specifications of the camera are given in Table 1. From the 
table it is seen that the CRC-1 is a large format (23 x 25 cm) roll film 
camera. Film is the chosen medium over glass plates for a number of reasons: 
(1) film is over 20 times less expensive per frame than commercially 
available plates of the highest degree of flatness (namely, 6 mm 
thick Kodak Microflat plates) and is much more readily available; 
(2) film is considerably easier to use and permits photographs to 
be taken far more rapidly and efficiently; 
(3) film can be rendered considerably flatter than Kodak Microflat 
plates by means of a properly designed vacuum platen; 
(4) by virtue of GSI's proprietary reseau platen (U.S. Patent 
No. 4,149,788) appropriate corrections for film deformation 
can readily be generated, thus reducing this source of error 
to a negligible level. 
(5) as will be explained presently, film can accomodate an 
indefinitely large number of multiple exposures made feasible 
through use of targets fabricated from retroreflective sheeting. 
It is true that photographic plates have a decided advantage over film 
with respect to dimensional stability, an advantage that film can overcome 
only through corrections made possible by means of a reseau incorporated & 
into the camera (as is the case with the CRC-1). On the other hand, it is 
not generally appreciated how poor even the best of plates are with respect 
to flatness. For example, with 24 by 24 cm Kodak Micro-Flat plates the rms 
departure from a best fitting plane averages about 7um with systematic 
excursions often exceeding 15um in the corners of the format. Plates of 
lower grades of flatness (e.g., Kodak Ultra-Flat or Gevaert Ultra-Plan) are 
several times worse than this. In most photogrammetric applications, errors 
attributable to unflatness are the dominant errors in results generated by 
plate cameras. This is best appreciated when one considers that x,y coor- 
dinates of well-defined images can be measured to an rms accuracy on the 
order of 2um, whereas errors attributable to unflatness of plates in 
cameras of moderate angular field can easily exceed the equivalent of 10um 
near the corners of the format. As is pointed out in Brown (1980), it is 
both possible and practical by means of air gauging to measure the surface 
70