14. 
The readings made by the human operator were made under somewhat 
adverse conditions. First of all, the flying spot on the surface of the cathode-ray 
tube that does the electronic scanning isused as a measuring mark. For measure- 
ment by an operator the motion of the spot may be stopped and then the spot is 
seen as a small bright dot in the center of the cathode-ray tube. This dot, however, 
is too bright to be an ideal measuring mark. Secondly, this mark was observed in 
an anaglyph system that is, naturally, affected by the requirements set forth by 
the automatic stereoplotting device. However, the measurements were carried 
out without major difficulties, and from the spread of repeated observations it was 
possible to find that the mean square error of one observation was better than 
1.0 meters. 
The most important part of the investigations on the accuracy of spot 
elevations was the comparison with elevations measured in the field. This mate- 
rial was processed in exactly the same manner as the other results of the Second 
International Mapping Experiment* and therefore the results are comparable. The 
elevation readings were corrected a posteriori for earth curvature and for residual 
model deformation. The final mean square error was: 
mp = Tl.8 meters. 
This compares quite well with the results obtained in the International Tests using 
instruments of the Kelsh Plotter type (mp = £1.33 m on the average). 
The accuracy of the human operator on the same model was slightly 
lower, 
my, = 12.2 meters. 
This is undoubtedly due partly to the adverse conditions mentioned above. 
It appears that there is a slight element of common error in both these 
determinations. This can be suspected on the basis of a comparison of the results 
of automatic stereoplotting device against the readings made by the human operator. 
The mean square value of the discrepancies is 12.5 meters, whereas a value to 
be expected on the basis of the two mean square errors mentioned above is 12.8 
meters. This may be interpreted to indicate that the accuracy of the automatic 
stereoplotting device may be better than shown above, although the evidence is too 
scanty to permit firm conclusions. 
The scanning páttern of the automatic stereo device is what is called a 
"random scan'". The flying spot on the surface of the cathode-ray tube deviates 
from its central position in a random manner in both x and y directions. Conse- 
quently, small deviations are more frequent than large ones and the gravity point 
oí the figure described by the scanning spot during any reasonable length of time 
coincides with the center point. This means that during the measurement the 
instrument is continuously weighing the information and it extracts it in such a 
way that more emphasis is placed on the information obtained in the immediate 
vicinity of the point that coincides with the rest position of the dot. Therefore, it 
might be possible to improve the resolving power of the device by increasing the 
relative frequency of small deviations from the center position. 
  
* Blachut, T.J. Second International Mapping Experiment — Renfrew Test Area. 
Results of experimental plotting for 1:50,000 maps. Canadian 
Surveyor, XV, No 3, 1960, 138-160.