Information assessment for polarimetric SAR intensity images

摘要

Polarimetric synthetic aperture radar (SAR) systems generate data, which have different information content in each polarimetric band. In analyzing the intensity information of multi-polarimetric SAR images (without taking into account phase information) two key issues were addressed. First, the mutual information and the entropy are used to assess the information difference between the polarimetric bands. Second, the image restoration method for full polarimetric SAR images can be evaluated with the information theory and a statistical method of minimum mean squared error estimation. This method was used to find the amount of information in an image as a function of spatial frequencies. A priori information about the power spectrum of an image and a point spread function of SAR system was applied. Statistical quality measures as well as the amount of information show good performances for polarimetric image restoration.