Study of adaptive multi-band polarization image fusion

To remove the clutter efficiently and detect the objects accurately, the information presented by multi-band polarization images should be combined. An adaptive weighting multi-band polarization images fusion method is proposed based on the false color mapping and the degree of linear polarization entropy to combine the information. The characters of object and background and the principles of multi-band imaging polarization are used in this method. At the same time, the relation between objects' spectral and polarization information, the information difference contained in polarization images in different band, the relations among Stokes' images in same band, and the information relation between object and background in the degree of linear polarization (DoLP) image are all used in this method. The background clutter is removed through combing Stokes' images and DoLP image in different band. In final simulation, this fusion method is used in different multi-band polarization images, the fusion results and evaluation index show that the background is removed efficiently.