Classification of a reference image using auxiliary images

This work does an extension of the Maximum A Posteriori (MAP) multidimensional image classification for the case that each image in a set of images can have different number of predominant classes. These different classes can be related to the sensor characteristics, which can detect different aspects of the same scene. In the proposed classification processes one image of the image set is choose as the reference image to be classified and the others images are considered complementary information. Simulations of two SAR amplitude images are used to evaluate the classification method.