238 
SYMPOSIUM PHOTO INTERPRETATION, DELFT 1962 
cost is only a small part of the total cost of aerial photography. When aircraft 
costs, and stand-by time for the flight crew, elimination of the need for prints 
and the ease of interpreting color film in rolls are considered, the extra cost of 
the color film is minor. Increased interpretation accuracy on color film would 
counter balance any slight increase in cost. 
Even though the color film used in this test was slightly overexposed, results 
showed it to be superior to panchromatic for tree species identification. 
In using Munsell color charts it was found that photo scale influenced the 
hue that interpreters designated as being the same species. For example, when 
these charts were used trees were called greener at the small scale and yellower 
at the larger scales. 
Tree species were identified accurately enough on the large scales of color 
film to suggest the possibility of using them on actual inventory problems. 
Further study of hardwoods, especially to associate their crown and foliage 
characteristics with physiographic features, should be helpful in the more 
accurate identification of these tree species. 
Discussion 
Mr. Le Ray (France) asked what method was used to measure the accuracy of the inter 
preters. Answer: The analysis of variance. 
Mr. Heller answered Dr. D. A. Francis (U.K./FAO) on a relevant question that the costs 
of large scale colour photography were five times as high as those of panchromatic photo 
graphy. Asked about the operational difficulties the speaker said that flying was carried out 
in the morning in order to avoid too much turbulence. 
Mr. Hempenius (Netherlands) asked if the time needed for the interpretation of colour film 
was less than for panchromatic film. Answer: Interpretation of the characteristics is quicker 
on colour than on black and white. 
Dr. G. Hildebrandt (B.R.D.) asked if comparable tests were carried out with smaller scales, 
e.g. 1 : 15,000. What were the results? Answer: Smaller scales gave no acceptable results for 
forest insect damage. The method used is a sampling method and is not meant for mapping 
purposes.