A knowledge based classification of multitemporal ERS-1 and JERS-1 SAR images over the tropics

Presents a knowledge based classification of multitemporal SAR images for land-cover application in general and agricultural applications in particular, by adapting methodologies more commonly applied to digital segmentation. The knowledge requirements are obtained from the temporal signature variations of the various land-cover types present in the area, with signature curves being based on ether radiometry or texture of the objects concerned. Use of textural information is made only in those cases when the use of object radiometry proves to be inadequate. Results show that classification accuracies are close to that obtained from maximum likelihood classification for the landsat TM data of the same area