Downloads: 5 | Views: 41 | Weekly Hits: ⮙1 | Monthly Hits: ⮙1

Research Paper | Geoinformatics | India | Volume 12 Issue 8, August 2023

Temporal Dual Polarization SAR Data for Multi Crop Classification Using Different Classifiers

Gopi P | Manjunath K. R

Abstract: The curiosity in crop inventory since beginning was the use of multitemporal microwave data so as to achieve a near perfect crop discrimination. Till recently, the multi - temporal amplitude data has been used for crop discrimination as well as biomass estimation. With the availability of dual - polarized data, the disparity response of crop geometry at different crop growth stages to different polarizations is being exploited for discrimination and classification of crops. The crop signature analysis has been made for Synthetic Aperture Radar (SAR) backscatter coefficient (?0) of paddy crop shows significant sequential behaviour and a large range during its growth period, which is due to the interaction of microwave energy with the crop canopy, increasing from the transplantion stage to the reproductive stage. This temporal variation of SAR backscatter clearly differentiates paddy fields from other land cover classes and also various crop signature shows unique pattern. The different classifiers were performed in this study site and comparison study was also carried out. The decision tree classifier based on the temporal assessment of SAR backscatter was attempted to classify crops and land use features, showed more than 84% accuracy. An attempt has been made in the current study using RISAT1 C - band dual, polarized data for crop inventory. The three date dual polarized data gave comparable results for crops like Green gram, Cluster beans, Ladies Finger, Paddy, Cotton, Fallow, water body and settlement.

Keywords: Remote Sensing, Synthetic Aperture Radar (SAR), Crop Discrimination

Edition: Volume 12 Issue 8, August 2023,

Pages: 641 - 648