Classification of Agricultural Areas in Multi-temporal TerraSAR-X Images using Morphological Profiles

Due to human population growth and global climate change, crop monitoring is becoming more significant every day. Remote sensing is a very practical method for crop monitoring. In this study, we classified agricultural crops in multi-temporal single-pol TerraSAR-X images. With the extraction of morphological profiles, spatial information is fused with SAR backscattering. K-NN algorithm is applied and overall classification accuracy in multi-temporal images reached 90% for 18 classes with the contribution of morphological profiles.