You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Transactions of the Symposium on Photo Interpretation

with involved mathematical computations. This is an interesting and a valuable
use of computer capabilities, but, however, a limited one in that the computer
is still being used only as a sort of extremely rapid calculating machine. There
are quite a number of other things that a computer can do, and it is the
purpose of this paper to describe a number of them as they apply specifically
to photo interpretation.
Perhaps the best first question is “Why do we need computers in photo
interpretation?” Such devices tend to be large, complex, and expensive, and
should not, therefore, be considered unless there is a problem situation where
in their capabilities provide a clear-cut, almost unique solution. I suggest
there are such problem areas in photo interpretation. I propose to discuss two
of them and then to report on the present status of applicable research at the
System Development Corporation.
The emerging problem areas
The problems involved can be summed up in two words: data load. This
data load problem has two distinct aspects: 1. a tremendously increased
volume of photography; and, 2. a resulting greatly expanded volume oi ex
tracted information. It will be valuable to examine for a moment some evidence
relating to the existence of problems in these areas.
I am sure that you are all aware of the tiros program which is now in
operation. Tiros is a weather reconnaissance satellite which orbits the earth
and sends back photographs of cloud type, distribution, storm activity, etc.
from which meteorologists can make weather forecasts of heretofore unequalled
accuracy and completeness. Table 1 provides an overview of the data load
problem with respect to tiros [1, 2, 3]. All three of these satellites represent:
experiments in what kind of equipment should be used, and how much photo-
Table 1. Weather satellite data load
Lens type
Time operating
Number of pictures
Wide and Narrow
Angle Lenses
78 days
Wide and Narrow
Angle Lenses
10 months
Wide Angle
60 days
graphy should be sent back to earth. It is, therefore, very difficult to attempt
generalization on the basis of the returns as we know them so far. However,
it seems that something like 12,000 pictures a month is not an unreasonable
estimate concerning the nature of the data load which should be expected.
The photo returns from the tiros system may, however, be actually a drop
in the bucket when compared to the volume of photography which could
arise out of an aerial reconnaissance system used in support of a world arms