TOWARD AUTOMATIC IMAGE CLASSIFICATION APPROACH FOR REMOTELY SENSED DATA

D. A. Algarni

King Saud University

Civil Engineering Department

P.O.Box 800, Riyadh 11421

Saudi Arabia

 

ABSTRACT

The purpose of this research is to seek an approach through which image features can be automatically recognized. The concept of image classification of remote sensing was applied but in a different way. The concentration was on mainly the qualitative parameters rather than on only that of the quantitative ones. The image features identification task is decomposed to smaller tasks, each of which has its specific function to perform at the appropriate time. Some of the sub-tasks are data extraction, hypotheses generation, and hypotheses evaluation. The experiment showed the possibility of applying a reasoning techniques such as that of the Generic Task (GT) approach to the task of image classification and interpretation.

 

 

ACSM/ASPRS Annual Convention & Exposition. Baltimore, USA, 1994. 2:219-227.