adjust

Author: yxdragon

imagepy/core/util/fileio.py

@@ -8,7 +8,6 @@
 def show_img(img, title):
     if isinstance(img, list):
         return IPy.show_img(img, title)
-    # if img.dtype!=np.uint8 and img.ndim>2 and img.shape[2]!=3:
     if img.ndim>2 and img.shape[2]!=3:
         return IPy.show_img(img, title)
     if img.dtype==np.uint8 and img.ndim==3 and img.shape[2]==4:

imagepy/core/wraper/imageplus.py

@@ -178,7 +178,6 @@ def lookup(self, img=None):
 
 
     def swap(self):
-        print(type(self.snap), type(self.imgs[self.cur]), self.cur)
         if self.snap is None:return
         if isinstance(self.imgs, list):
             self.snap, self.imgs[self.cur] = self.imgs[self.cur], self.snap

imagepy/menus/Image/Adjust/enhanceContrast_plg.py

@@ -7,11 +7,11 @@
 import numpy as np
 
 class Plugin(Filter):
-    title = 'Enhance contrast'
+    title = 'Enhance Contrast'
     note = ['all', 'auto_msk', 'auto_snap','preview']
-    para = {'percentage': 0.3}
-    view = [(float, 'percentage', (0,100), 4, 'Saturated pixels', '%')]
+    para = {'percentage': 5}
+    view = [(float, 'percentage', (0,100), 2, 'Saturated pixels', '%')]
 
     def run(self, ips, snap, img, para = None):
-        up, down = np.percentile(snap, (0, 100 - para['percentage']))
+        up, down = np.percentile(snap, (para['percentage']/2, 100-para['percentage']/2))
         return exposure.rescale_intensity(snap, in_range=(up, down))

imagepy/menus/Image/Adjust/normalize_plg.py

@@ -8,26 +8,27 @@
 
 class Plugin(Simple):
     title = 'Normalize'
-    note = ['8-bit','16-bit','float','stack']
-    para = {'if3d': False, 'Sb':False}
-    view = [(bool, 'if3d', '3D stack'),
-    (bool, 'Sb', 'Subtract background')]
+    note = ['8-bit','16-bit','float']
+    para = {'is3d': False, 'sb':True}
+    view = [(bool, 'is3d', '3D stack'),
+            (bool, 'sb', 'Subtract background')]
 
     def run(self, ips, imgs, para = None):
-        imgs_= np.array(imgs).astype('float64').transpose(1,2,0)
-        if para['if3d']:
-            if para['Sb']:
-                imgs_ -= imgs_.min()
-            imgs_ = imgs_ / imgs_.max()
-        else:
-            if para['Sb']:
-                for i in range(z):
-                    imgs_[:,:,i] -= imgs_[:,:,i].min()
-            for i in range(z):
-                imgs_[:,:,i] = imgs_[:,:,i] / imgs_[:,:,i].max()        
-        if imgs.dtype == np.uint8:
-            imgs[:] = (255*imgs_).astype(imgs.dtype).transpose(2,0,1)
-        elif imgs.dtype == np.uint16:
-            imgs[:] = (65535*imgs_).astype(imgs.dtype).transpose(2,0,1)
-        else:
-            imgs[:] = (imgs_).astype(imgs.dtype).transpose(2,0,1)
+        lim = np.zeros([len(imgs), 2], dtype=imgs[0].dtype)
+        dic = {np.uint8:255, np.uint16:65535, np.float32:1, np.float64:1}
+
+        IPy.set_info('count range ...')
+        for i in range(len(imgs)):
+            lim[i] = imgs[i].min(), imgs[i].max()
+            self.progress(i, len(imgs))
+        
+        maxvalue = dic[imgs[0].dtype.type]
+        if not para['sb']: lim[:,0] = 0
+        rg = lim[:,0].min(), lim[:,1].max()
+        if para['is3d']: lim[:] = rg
+        IPy.set_info('adjust range ...')
+        for i in range(len(imgs)):
+            if para['sb']: imgs[i] -= lim[i,0]
+            np.multiply(imgs[i], maxvalue/(lim[i].ptp()), out=imgs[i], casting='unsafe')
+            self.progress(i, len(imgs))
+        ips.range = 0, maxvalue