Preserve image histogram range, when loading, processing and saving.

Update cli interface to simplify and support PixInsight script integration
Add PyxInsight cli script

.gitignore

@@ -18,3 +18,6 @@ models/**
 *.tif*
 *.zip
 *.tfevents.*
+pjsr/
+safe/
+updates/

astrodenoise/applayout.py

@@ -296,7 +296,7 @@ def save(self, path):
             self.dismiss_popup()
 
         filepath = Path(path)
-        #AstroDeNoiseApp()
+
         get_app(App.get_running_app()).lastpath = str(filepath.parent)
 
         if self.fits_headers is None:
@@ -317,14 +317,19 @@ def save(self, path):
         else:
             result_tosave = np.array(self.currentimage.get_data('pre')[0])
 
+        # Clip to range [0,1] before save
+        result_tosave = np.clip(result_tosave, 0, 1)
+
         try:
             extension = filepath.suffix.lower()
 
             if extension in supported_save_formats_fits:
-                result_forsave = np.moveaxis(np.transpose(result_tosave),1,2)
-                write_fits(filepath,result_forsave,headers=self.fits_headers)
+                write_fits(
+                    filepath,
+                    np.moveaxis(np.transpose(result_tosave),1,2),
+                    headers=self.fits_headers)
             elif extension in supported_save_formats_tiff:
-                result_tosave = (result_tosave - np.min(result_tosave)) / (np.max(result_tosave) - np.min(result_tosave))
+                # Scale to unit16 for tiff
                 result_tosave = (result_tosave * np.iinfo(np.uint16).max).astype(np.uint16)
                 imsave(filepath,data=result_tosave)
             else:
@@ -459,7 +464,11 @@ def denoise(self, data, C=-2.8,B=0.25):
 
         self.update_progress(0)
         expand_low_actual = 0.5 - (self.expand_low/2)
-        normalizer = STFNormalizer(C=C,B=B,expand_low=expand_low_actual,do_after=False) if self.normalize_enabled else NoNormalizer(expand_low=expand_low_actual)
+        # Strength shifts pixel values to right of histogram by up to 0.5
+        #Strength=1 => Image + 0
+        #Strength=0 => Image + 0.5
+        normalizer = STFNormalizer(C=C,B=B,expand_low=expand_low_actual,do_after=False) if self.normalize_enabled else NoNormalizer(expand_low=expand_low_actual,do_after=False)
+
 
         if self.denoise_enabled:
             with tf.device(f"/{self.selected_device}:0"):
@@ -478,7 +487,7 @@ def denoise(self, data, C=-2.8,B=0.25):
             result = normalizer.before(np.moveaxis(np.transpose(data),0,1),'YX')
             self.update_progress(1)
 
-        return result 
+        return result - expand_low_actual
 
     @mainthread
     def update_progress(self,progress):
@@ -526,8 +535,9 @@ def get_label_data(self):
         }
 
     def get_texture(self, result):
-        image = (result - np.min(result)) / (np.max(result) - np.min(result))
-        image = (image * 255).astype('uint8')
+
+        result = np.clip(result, 0, 1)
+        image = (result * 255).astype('uint8')
 
         colorfmt='rgb'
         if image.shape[2] == 1:            

astrodenoise/cli.py

@@ -1,7 +1,9 @@
 import os
+import sys
 import argparse
 import tensorflow as tf
 from tifffile import imread
+from xisf import XISF
 import numpy as npp
 from pathlib import Path
 from os.path import join as path_join
@@ -11,20 +13,27 @@
 from astrodeep.utils.fits import read_fits, write_fits
 from astrodenoise.version import modelversion
 
+def get_exepath():
+    if getattr(sys, "frozen", False):
+        datadir = os.path.dirname(sys.executable)
+    else:
+        datadir = Path(os.path.dirname(__file__)).parent.as_posix()
+    return datadir
+
 def cli():
 
     parser = argparse.ArgumentParser()
 
     parser.add_argument('input', type=str, nargs=1, help='Input image path, either tif or debayered fits file with data stored as 32bit float.')
     parser.add_argument('--model','-m', type=str, default=modelversion, help='Alternative model name to use for de-noising.')
     parser.add_argument('--models_folder', type=str, default='models', help='Alternative models folder root path.')
-    parser.add_argument('--tiles','-t', type=int, default=0, help='Use number of tiling slices when de-noising, useful for large images and limited memory.')
+    parser.add_argument('--tiles','-t', type=int, default=3, help='Use number of tiling slices when de-noising, useful for large images and limited memory.')
     parser.add_argument('--overwrite','-o', action='store_true', help='Allow overwrite of existing output file. Default: False when not specified.')
     parser.add_argument('--device','-d', choices=['GPU','CPU'], default='CPU', help='Optional select processing to target CPU or GCP. Default: CPU')
     parser.add_argument('--normalize','-n', action='store_true', help='Enable STFNormalization before de-noising. Default: False when not specified.')
     parser.add_argument('--norm-C', type=float, default=-2.8, help='C parameter for STF Normalization. Default: -2.8')
-    parser.add_argument('--norm-B', type=float, default=0.25, help='B parameter for STF Normalization, Higher B results in stronger stretch providing the ability target de-noising more effectively. . Default: 0.25, Range: 0 < B < 1')
-    parser.add_argument('--norm-restore', action='store_true', help='Restores output image to original data range after processing. Default: False when not specified.')
+    parser.add_argument('--norm-B', type=float, default=0.25, help='B parameter for STF Normalization, Higher B results in stronger stretch providing the ability target de-noising more effectively. . Default: 0.25, Range: 0 < B < 1')    
+    parser.add_argument('--strength', type=float, default=0.5, help='The denoise strength applied. Default: 0.5')
 
     args = parser.parse_args()
 
@@ -35,7 +44,9 @@ def predict(path,model):
             if path.suffix in ['.fit','.fits']:
                 data, headers = read_fits(path)
             elif path.suffix in ['.tif','.tiff']:
-                data, headers = npp.moveaxis(imread(path),-1,0), None            
+                data, headers = npp.moveaxis(imread(path),-1,0), None   
+            elif path.suffix in ['.xisf']:
+                data, headers = npp.moveaxis(XISF(path).read_image(0),-1,0), None
             else:
                 print("Skipping unsupported format. Allowed formats: .tiff/.tif/.fits/.fit")
                 return
@@ -46,33 +57,38 @@ def predict(path,model):
             if data.ndim == 2:
                 data = data[npp.newaxis,...]
 
-            print("Processing file:",path)
-            print("Image Dimensions:",data.shape)
+            print(f"Processing file:{path}\n")
+            print(f"Image Dimensions: {data.shape}\n")
 
-            n_tiles = None if args.tiles == 0 else (args.tiles,args.tiles)
+            n_tiles = None if args.tiles == 0 else (args.tiles, args.tiles)
             if n_tiles is not None:
                 print("Processing with tilling:",n_tiles)
 
-            output_denoised = []
+
             axes = 'YX'
-            normalizer = STFNormalizer(C=args.norm_C,B=args.norm_B,do_after=args.norm_restore) if args.normalize is True else NoNormalizer()
-            print("Using Normalization:",normalizer.params)
+            expand_low_actual = 0.5 - (args.strength/2)
+            normalizer = STFNormalizer(C=args.norm_C,B=args.norm_B,expand_low=expand_low_actual,do_after=True) if args.normalize is True else NoNormalizer(expand_low=expand_low_actual,do_after=True)
+            print(f"Using Normalization: {normalizer.params}\n")
+            print(f"Using Strength: {args.strength}\n")
 
+            output_denoised = []
             for c in data:
                 output_denoised.append(
-                    model.predict(c, axes, normalizer=normalizer,resizer=PadAndCropResizer(), n_tiles=n_tiles)
+                    model.predict(c, axes, normalizer=normalizer, resizer=PadAndCropResizer(), n_tiles=n_tiles)
                     )
 
+            output_denoised_arr = npp.asarray(output_denoised)
+            # Clip to [0,1] range            
+            output = output_denoised_arr.clip(0,1)
+
             output_file_name = path.stem + f"_denoised.fits"
-            output_path = path_join(path.parent, 'denoised')
-            Path(output_path).mkdir(exist_ok=True)
-            output_file_path = path_join(output_path, output_file_name)
-            write_fits(output_file_path, output_denoised, headers, args.overwrite)
-            print("Output file saved:", output_file_path)
+            output_file_path = path_join(path.parent, output_file_name)
+            write_fits(output_file_path, output, headers, args.overwrite)
 
+            print("Output file saved:", output_file_path)
 
         print("Loading model:", args.model)
-        model = CARE(config=None, name=args.model, basedir=args.models_folder)    
+        model = CARE(config=None, name=args.model, basedir=Path(get_exepath()).joinpath(args.models_folder).as_posix())
         file_or_path = args.input[0]
 
         if os.path.isfile(file_or_path):