Fusion of Evidences for Edge Detection in PolSAR Images

Polarimetric Synthetic Aperture Radar (PolSAR) has achieved an important position as a remote sensing imaging method. However, PolSAR images are contaminated with speckle noise, making its processing and analysis challenging tasks. The present study discusses a detection method based on the fusion of evidences obtained in the intensity channels of multilook PolSAR images. The method consists of detecting transition points in the finest strip of data which spans two regions using the maximum likelihood. This is applied to each of the three intensity channels (hh), (hv) and (vv). The fusion methods are simple average, stationary wavelet transform (SWT), principal component analysis (PCA), and ROC statistics. The results indicate improvement performance of the approach in detecting edges with possible paths for future research.