Novel Features for Polarimetric SAR Image Classification by Neural Network

This paper presents a set of effective features derived from the coherence matrix of polarimetric SAR data. Neural network is used as the classification engine. The maximum likelihood estimator (MLE) result is used as the reference to compare the result of the proposed method. It is demonstrated that the average classification accuracy by the proposed method is more than that by the MLE. The maximum overall efficiency obtained by the proposed method is 95.4%.