Wetland cover information extraction research based on the multi-polar radar images and multi-spectrum optical images fusion

The research takes Radarsat-2 multi-polarization radar images and ALOS multi-spectrum optical images for example, through the information filtering and extraction methods such as information quantity statistics and correlation matrix, we obtain the cognition of the spectral characteristics and detection performance of each polarization mode radar image, and select the optimal data for fusion and classification experiment. Using Brovey, PCA, HPF and Wavelet transformation methods to fuse, and adopting non-supervision, supervision and object-oriented method to carry out classification experiment, we explore the application of radar and optical remote sensing image fusion technology in the wetland area covers information extraction, which verifies the effectiveness and feasibility using radar and multi-spectral image fusion technology in the wetland cover information extraction.