Crop classification from airborne synthetic aperture radar data

The ability of radar systems to record useful data independently of the prevailing weather conditions and the future increase in radar data availability is likely to result in their increased utilization for crop classification. However, a variety of factors relating to the quantity and quality of the data will influence the accuracy with which these data can be classified. This paper aims to illustrate the effect of some of these factors on crop classification accuracy from a multi-feature synthetic aperture radar data base. The results show that in addition to the number of data channels available for a classification, the method of radiometric correction applied to the data and the use of data from different look directions can have significant effects on the classification accuracy. The use of a multi-feature data base however, was found to enable accurate discrimination of a variety of crop types.