I`m triying to compare some supervised classifications and I need to use the rule images for the maximum likelihood classification. I cannot use these images because they have values ranging from -300.000 up to 100. I have tried to change the scale factor several times, but I cannot obtain any reasonable values... What I need is a probability image ranging from 0-1 or any other proportional variation. Can anyone help me?, With the rule classifier I only obtain binary images (0 or 1 values), but not the probability images. Thanks