Development of an assisting system for the integration of SAR and InSAR images for the Belgian Ministry of Agriculture

A SAR images preprocessing chain has been adapted to be incorporated in the crop control system of the Belgian Ministry of Agriculture. The use of SAR images in addition to the available optical images increases the efficiency of the fraudulently declared parcels detection. The classification waterfall strategy developed in this project allows to increase the efficiency of the control for any remote sensing data. Indeed this is improved the control when achieved by optical images as well through the combination of optical an radar images. A computer-assisted photo interpretation method dedicated to the SAR data and adapted to the environment of the Ministry has been succesfully proposed.
Based on the results obtained for the year 2000, some conclusions can be translated in practical recommendations:

the use of 3 to 5 SAR images with 2 or 3 optical images to increase the crops (an crop groups) discrimination in comparison with the control system based only on optical data;

the use of the classification waterfall strategy to improve the control system efficiency based on both, optical data only and optical-SAR combination;

the substitution of the last optical image by 3 RADARSAT images in order to reduce by more than 1.5 month the time delivery of the diagnostic and maintain the efficiency level thanks to the waterfall strategy;

the use of SAR data for the photo interpretation of the suspected parcels using successive filtering techniques and the relative and local comparison protocol.