CROP CHANGE ASSESSMENT USING POLARIMETRIC RADARSAT-2 DATA

摘要

This paper studies the feasibility of monitoring croprngrowth cycles based on a temporal variation analysis ofrnthree elementary radar scattering mechanisms. Croprnchanges are assessed using RADARSAT-2 polarimetricrndata. The polarimetric SAR (PolSAR) analysis is basedrnon the Pauli decomposition. Multi-temporal analysis isrnapplied to RGB images constructed using surfacernscattering, double bounce and volume scattering. Therncrops studied in this paper are corn, cereals andrnsoybeans. Each crop has unique physical structuralrncharacteristics and responds differently to thesernscattering mechanisms. By monitoring the significantrnchanges that occur in these scattering mechanisms, therncrop growth to harvest cycle can be observed and thernharvest time can be estimated. In addition, a MaximumrnLikelihood Classification was performed on thernRADARSAT-2 data to produce a crop map. An overallrnclassification accuracy of 85% was achieved.