/5 lens 
itive to 
would 
stop of 
that at 
werage 
0.4 the 
  
  
duces in a 
ed for all 
ll be found 
wiation of 
is not sig- 
y to use a 
dardize on 
lacing the 
ope which 
gles, teles- 
or glasses 
ass it may 
he camera 
ariation in 
as a two- 
ariation in 
1 the same 
tion vector 
Of course, 
iffer from 
reference 
easily ac- 
tating the 
ent. Alter- 
beam may 
once up 
or sim- 
ted and 
xl. For 
4 times 
5 but 
  
be stated. 
sstablished 
CAMERA CALIBRATION CONSTANTS 
d. A, Eden 
Erm : uet 
Précis 
eme ares 
It is advocated that whatever information is provided on a 
calibration certificate the basic measures of x and © should always be 
included. 
  
e 
2 
The fundamental constants of a camera may be expressed by listing 
the external angles of entry © against the measures on the plate to 
which they apply. This is the basis of all camera calibration. Why 
is it then that these measures are not listed on the calibration 
certificate? 
The usual practice is to state on the calibration certificate a 
value for the principal distance against & list of distortions corres- 
© 
mh 
ponding to various angular distances from a point of origin. These 
are not basic measures at all. They are obtained from the original 
measures d a mathematical process involving various adjustments, 
distributions, interpolations and weightings. It is well known that 
these dis Stortion: may be expressed in & large variety of different 
Ways. Furthermore, the distortions are often given in the form of a 
graph in which the number of observations made and the degree of inter- 
 Dolation employed is effectively hidden. 
The writer believes that the analysis of the basic measures should 
be undertaken by the photogrammetrist using a method which suits best 
his own purposes, and that there is no need for the analysis of these 
measures to be undertaken by the calibrating authority. If this 
principle was adopted it would altogether absolve the calibrating 
authority from the necessity of explaining his computing techniques. 
Under present practice if he does not do this in sufficient detail it 
is impo ssible for the photogrammetrist to work back from the data 
supplied to determine correctly the original observations. Thess it 
is  Sosentind for him to know if he wishes to determine his own 
distortions by & method most suitable for the task he has in mind. 
The writer therefore advocates that whatever information is 
)jvVided on a calibration certificate, the basic measures of x and 
a. 
qo 
should always be includec 
m liae