adapt M and m depending on model

deepdespeckling/merlin/merlin_denoiser.py

@@ -165,9 +165,9 @@ def preprocess_denoised_image(self, denoised_image_real_part: np.array, denoised
             processed denoised full image, processed denoised image real part, processed denoised image imaginary part
         """
         denoised_image_real_part = denormalize_sar_image(
-            denoised_image_real_part / count_image)
+            denoised_image_real_part / count_image, self.M, self.m)
         denoised_image_imaginary_part = denormalize_sar_image(
-            denoised_image_imaginary_part / count_image)
+            denoised_image_imaginary_part / count_image, self.M, self.m)
 
         # combine the two estimation
         output_clean_image = 0.5 * (np.square(

deepdespeckling/utils/utils.py

@@ -22,11 +22,13 @@ def normalize_sar_image(image: np.array) -> np.array:
     return ((np.log(np.clip(image, 0, image.max())+1e-6)-m)/(M-m)).astype(np.float32)
 
 
-def denormalize_sar_image(image: np.array) -> np.array:
+def denormalize_sar_image(image: np.array, M: float = M, m: float = m) -> np.array:
     """Denormalize a sar image store in  a numpy array
 
     Args:
         image (numpy array): a sar image
+        M (float): global maximum used for normalisation
+        m (float): global minimum used for normalisation
 
     Raises:
         TypeError: raise an error if the image file is not a numpy array