The requirements of the evaluation product, the available time, personal and money are 
also to be considered. The product type, e.g. map, digital data, statistics, can favour 
a technique, which produces this type very economically, even when the accuracy or 
ease of the evaluation per se may be better with another procedure. Also the aim of the 
evaluation will determine the best procedure. A general evaluation to select small areas 
for detailed investigation may require another solution than a detailed mapping project. 
To select the most suitable procedure for a specific task, the requirements of the 
application and the possibilities of the evaluation techniques should match as close as 
possible. This procedure should also meet the needs with the smallest effort possible. 
EXAMPLES 
Two examples are used to demonstrate the potentials and requirements of the different 
techniques: 
1. Land use evaluation with LANDSAT data 
2. Selection of areas for detailed geological investigations. 
The actual land use is an important information in planning decisions. Interpretation of 
aerial photographs proved to be very useful, but for vast areas the amount of money and 
time spent can become overwhelming. Therefore the use of LANDSAT data seemed to 
offer a solution. It is evident that LANDSAT data can not yield results, which are as 
detailed as with aerial photographs. Therefore only limited requirements can be met, 
e.g. a positional accuray of 50 - 100 m, and the differentiation of about 10 to 15 classes. 
But the 3 techniques offer strikingly different potentials. 
An experienced photointerpreter can immediately recognize the main land use classes in 
a simulated color infrared rendition of the LANDSAT data, without specific ground 
information for that scene. The accuracy and detail will be determined by the image 
quality and his experience with comparable data within the same climatic region. Since 
he has an overview of the area, he may draw conclusions, regarding his problem, which 
go beyond the pure identification of objects. He can focus his attention on the areas or 
objects within the scene which are important for his decision making process. So he may 
be able to get the land use information he needs within a few minutes. But if he actually 
must map very detailed the different land use categories and digitize the results for 
storage in a digital data base, he has to work for days or weeks. 
6 
a 
  
Fi 
  
  
A si 
rule 
has 
of t 
for 
are 
val 
diff