25
Personnel training problems are more critical with this type of
equipment for two reasons. First, the operator must be familiar with
computer operation as well as the photogrammetric operations of the
instrument. Second, most efficient exploitation of this instrument
is achieved by a proper blending of numerical and analog techniques.
More than ever, an operator must be familiar with analytical photo
grammetric theory - a requirement formerly reserved only for key first
order instrument operators in a line organization.
To make maximum utilization of the Black Box Program it is
desirable to have an auxiliary computer system available to
compute the least squares solution for its coefficients. This involves
computer scheduling problems to make such computations in a timely
manner.
l4. Operations Performance Factors
With the experience acquired so far at ACIC, several statements
can be made about expected performance of the AP--2 in future operations.
As far as basic precision is concerned, the present AP-2
prototype does not equal the precision of,say the Wild A-7 • Witness
the AP-2 grid test precision of C ~ 5000 as opposed to C ~ 10,000
for an A-7 model magnification of 2:1. (The disparity is attributable
to the difference in model magnification scale.) This limitation on
basic precision is electronic in nature and can be improved by decreasing
the basic 5p increment to 2 l/2|_i off of the hand and foot wheels at
the expense of slew speed. On the other hand, when working with
average wide angle cartographic materials, the AP~2 performs as well
as the A-7. Consequently, it is our opinion that under normal
conditions, the AP-2 will perform with the same accuracy as present
available first order equipments.
In the comparison of operation times to conventional instrumenta
tion, the AP-2 does not show any significant advantages. What is
gained in the use of its automatic features in relative and absolute
orientation is lost in computer set up and tape preparation Also,
the need for the operator judgement is not eliminated with such
automation. Since the least squares process of relative orientation
pre-supposes a given math model, it is still the responsibility of the
operator to determine any deviations from the model and improve his
solution accordingly. If the Black Box solution is used, actual model
set up time may take longer than present procedures, but with an
accompanying improvement of accuracy.
As far as compilation time is concerned, no improvement is
expected since procedures on the AP-2 parallel what is presently
being done on conventional instruments.