Merge pull request #278 from wuan/tv_denoise_fix

fixed data of tv_denoise result images to be in default range for floats

skimage/filter/_tv_denoise.py

@@ -1,4 +1,5 @@
 import numpy as np
+from skimage import img_as_float
 
 
 def _tv_denoise_3d(im, weight=100, eps=2.e-4, n_iter_max=200):
@@ -26,7 +27,7 @@ def _tv_denoise_3d(im, weight=100, eps=2.e-4, n_iter_max=200):
     Returns
     -------
     out: ndarray
-        denoised array
+        denoised array of floats
 
     Notes
     -----
@@ -109,7 +110,7 @@ def _tv_denoise_2d(im, weight=50, eps=2.e-4, n_iter_max=200):
     Returns
     -------
     out: ndarray
-        denoised array
+        denoised array of floats
 
     Notes
     -----
@@ -169,9 +170,8 @@ def _tv_denoise_2d(im, weight=50, eps=2.e-4, n_iter_max=200):
                 E_previous = E
         i += 1
     return out
-
-
-def tv_denoise(im, weight=50, eps=2.e-4, keep_type=False, n_iter_max=200):
+    
+def tv_denoise(im, weight=50, eps=2.e-4, n_iter_max=200):
     """
     Perform total-variation denoising on 2-d and 3-d images
 
@@ -192,19 +192,13 @@ def tv_denoise(im, weight=50, eps=2.e-4, keep_type=False, n_iter_max=200):
 
             (E_(n-1) - E_n) < eps * E_0
 
-    keep_type: bool, optional (False)
-        whether the output has the same dtype as the input array.
-        keep_type is False by default, and the dtype of the output
-        is np.float
-
     n_iter_max: int, optional
         maximal number of iterations used for the optimization.
 
     Returns
     -------
     out: ndarray
-        denoised array
-
+        denoised array of floats
 
     Notes
     -----
@@ -245,7 +239,7 @@ def tv_denoise(im, weight=50, eps=2.e-4, keep_type=False, n_iter_max=200):
     """
     im_type = im.dtype
     if not im_type.kind == 'f':
-        im = im.astype(np.float)
+        im = img_as_float(im)
 
     if im.ndim == 2:
         out = _tv_denoise_2d(im, weight, eps, n_iter_max)
@@ -254,7 +248,4 @@ def tv_denoise(im, weight=50, eps=2.e-4, keep_type=False, n_iter_max=200):
     else:
         raise ValueError('only 2-d and 3-d images may be denoised with this '
                          'function')
-    if keep_type:
-        return out.astype(im_type)
-    else:
-        return out
+    return out      

skimage/filter/tests/test_tv_denoise.py

@@ -2,7 +2,7 @@ import numpy as np
 from numpy.testing import run_module_suite
 
 from skimage import filter, data, color
-from skimage import img_as_uint
+from skimage import img_as_uint, img_as_ubyte
 
 
 class TestTvDenoise():
@@ -27,12 +27,21 @@ class TestTvDenoise():
         grad_denoised = ndimage.morphological_gradient(
             denoised_lena, size=((3, 3)))
         # test if the total variation has decreased
+        assert grad_denoised.dtype == np.float
         assert (np.sqrt((grad_denoised**2).sum())
                 < np.sqrt((grad**2).sum()) / 2)
-        denoised_lena_int = filter.tv_denoise(img_as_uint(lena),
-                                              weight=60.0, keep_type=True)
-        assert denoised_lena_int.dtype is np.dtype('uint16')
 
+    def test_tv_denoise_float_result_range(self):
+        # lena image
+        lena = color.rgb2gray(data.lena())[:256, :256]
+        int_lena = np.multiply(lena, 255).astype(np.uint8)
+        assert np.max(int_lena) > 1       
+        denoised_int_lena = filter.tv_denoise(int_lena, weight=60.0)
+        # test if the value range of output float data is within [0.0:1.0]
+        assert denoised_int_lena.dtype == np.float
+        assert np.max(denoised_int_lena) <= 1.0
+        assert np.min(denoised_int_lena) >= 0.0        
+        
     def test_tv_denoise_3d(self):
         """
         Apply the TV denoising algorithm on a 3D image representing
@@ -45,13 +54,10 @@ class TestTvDenoise():
         mask += 20 * np.random.randn(*mask.shape)
         mask[mask < 0] = 0
         mask[mask > 255] = 255
-        res = filter.tv_denoise(mask.astype(np.uint8),
-                                weight=100, keep_type=True)
-        assert res.std() < mask.std()
-        assert res.dtype is np.dtype('uint8')
         res = filter.tv_denoise(mask.astype(np.uint8), weight=100)
-        assert res.std() < mask.std()
-        assert res.dtype is not np.dtype('uint8')
+        assert res.dtype == np.float
+        assert res.std() * 255 < mask.std() 
+
         # test wrong number of dimensions
         a = np.random.random((8, 8, 8, 8))
         try: