INTRODUCTION OF NEIGHBORHOOD INFORMATION IN EVIDENCE THEORY AND APPLICATION TO DATA FUSION OF RADAR AND OPTICAL-IMAGES WITH PARTIAL CLOUD COVER

Two ways of introducing spatial information in Dempster-Shafer evidenc e theory are examined: in the definition of the monosource mass functi ons, and, during data fusion. In the latter case, a ''neighborhood'' m ass function is derived from the label image and combined with the ''r adiometric'' masses, according to the Dempster orthogonal sum. The mai n advantage of such a combination law is to adapt the importance of ne ighborhood information to the level of radiometric missing information . The importance of introducing neighborhood information has been illu strated through the following application: forest area detection using radar and optical images showing a partial cloud cover. (C) 1998 Patt ern Recognition Society. Published by Elsevier Science Ltd. All rights reserved.