Research on decomposition and fusion of color and polarization image based on SCM

Abstract

The traditional fusion method for polarization image and color image is difficult to be satisfied in practice, so the method based on SCM in the fusion of polarization image and color image is designed and combined with HSV color field adaptive enhancement, so as to achieve the effect of distinguishing objects under shadow or reflection. Using the multi-scale decomposition of guided filter and Gaussian filter, different fusion schemes are proposed for different scale information: the average energy fusion method is used for the low-frequency coefficients, and the SCM with the edge energy (EOE) as the stimulation is used to replace the traditional high-frequency fusion rules, so as to enhance the spatial information expression. Finally, the parameters are corrected by the consistency test. Experiments show that this method not only has a good color expression ability, but also effectively adds the polarization information of the target object, and achieves good performance in objective and supervisor evaluation.