Fusion of POLSAR and Multispectral Satellite Images: A New Insight for Image Fusion

Polarized synthetic aperture radar (POLSAR) and multispectral images have great complementarity in information volume. That is to say, POLSAR have high resolution but poor color information. Multispectral images have rich spectral channel information, but the resolution is low. Therefore, this work has explored the fusion problem of the two data source. A framework was proposed, which merged polarized channel fusion data and multispectral images based on the Sentinel-2 and GF-3 data. The experimental results showed that the fusion results greatly integrated the characteristics of each channel of POLSAR and optical image. Therefore, our work has great application potential in improving the accuracy of feature recognition.