Merge pull request #1951 from jmetz/update_peak_local_max_exclude_border

Updated exclude_border to not use min_distance

skimage/feature/peak.py

@@ -24,17 +24,16 @@ def peak_local_max(image, min_distance=1, threshold_abs=None,
     min_distance : int, optional
         Minimum number of pixels separating peaks in a region of `2 *
         min_distance + 1` (i.e. peaks are separated by at least
-        `min_distance`). If `exclude_border` is True, this value also excludes
-        a border `min_distance` from the image boundary.
+        `min_distance`).
         To find the maximum number of peaks, use `min_distance=1`.
     threshold_abs : float, optional
         Minimum intensity of peaks. By default, the absolute threshold is
         the minimum intensity of the image.
     threshold_rel : float, optional
         Minimum intensity of peaks, calculated as `max(image) * threshold_rel`.
-    exclude_border : bool, optional
-        If True, `min_distance` excludes peaks from the border of the image as
-        well as from each other.
+    exclude_border : int, optional
+        If nonzero, `exclude_border` excludes peaks from
+        within `exclude_border`-pixels of the border of the image.
     indices : bool, optional
         If True, the output will be an array representing peak
         coordinates.  If False, the output will be a boolean array shaped as
@@ -89,11 +88,14 @@ def peak_local_max(image, min_distance=1, threshold_abs=None,
 
     >>> img2 = np.zeros((20, 20, 20))
     >>> img2[10, 10, 10] = 1
-    >>> peak_local_max(img2, exclude_border=False)
+    >>> peak_local_max(img2, exclude_border=0)
     array([[10, 10, 10]])
 
     """
 
+    if type(exclude_border) == bool:
+        exclude_border = min_distance if exclude_border else 0
+
     out = np.zeros_like(image, dtype=np.bool)
 
     # In the case of labels, recursively build and return an output
@@ -137,12 +139,12 @@ def peak_local_max(image, min_distance=1, threshold_abs=None,
         image_max = ndi.maximum_filter(image, size=size, mode='constant')
     mask = image == image_max
 
-    if exclude_border and (footprint is not None or min_distance > 0):
+    if exclude_border:
         # zero out the image borders
         for i in range(mask.ndim):
             mask = mask.swapaxes(0, i)
             remove = (footprint.shape[i] if footprint is not None
-                      else 2 * min_distance)
+                      else 2 * exclude_border)
             mask[:remove // 2] = mask[-remove // 2:] = False
             mask = mask.swapaxes(0, i)
 

skimage/feature/tests/test_peak.py

@@ -145,8 +145,34 @@ def test_ndarray_exclude_border():
     nd_image[2,2,2] = 1
     expected = np.zeros_like(nd_image, dtype=np.bool)
     expected[2,2,2] = True
-    result = peak.peak_local_max(nd_image, min_distance=2, indices=False)
-    assert (result == expected).all()
+    expectedNoBorder = nd_image > 0
+    result = peak.peak_local_max(nd_image, min_distance=2,
+        exclude_border=2, indices=False)
+    assert_equal(result, expected)
+    # Check that bools work as expected
+    assert_equal(
+        peak.peak_local_max(nd_image, min_distance=2,
+            exclude_border=2, indices=False),
+        peak.peak_local_max(nd_image, min_distance=2,
+            exclude_border=True, indices=False)
+        )
+    assert_equal(
+        peak.peak_local_max(nd_image, min_distance=2,
+            exclude_border=0, indices=False),
+        peak.peak_local_max(nd_image, min_distance=2,
+            exclude_border=False, indices=False)
+        )
+    # Check both versions with  no border
+    assert_equal(
+        peak.peak_local_max(nd_image, min_distance=2,
+            exclude_border=0, indices=False),
+        expectedNoBorder,
+        )
+    assert_equal(
+        peak.peak_local_max(nd_image,
+            exclude_border=False, indices=False),
+        expectedNoBorder,
+        )
 
 
 def test_empty():