Add documentation for rgb2hsv.

scikits/image/color/colorconv.py

@@ -14,9 +14,31 @@ __docformat__ = "restructuredtext en"
 import numpy as np
 
 def rgb2hsv(rgb):
+    """RGB to HSV color space conversion.
 
-    """
-    RGB to HSV color space conversion    
+    Parameters
+    ----------
+    rgb : ndarray
+        The image in RGB format, in a 3-D array of shape (.., .., 3).
+
+    Returns
+    -------
+    out : ndarray
+        The image in HSV format, in a 3-D array of shape (.., .., 3).
+
+    Raises
+    ------
+    ValueError
+        If `rgb` is not a 3-D array of shape (.., .., 3).
+
+    Examples
+    --------
+    >>> import os
+    >>> from scikits.image import data_dir
+    >>> from scikits.image.io import imread
+
+    >>> lena = imread(os.path.join(data_dir, 'lena.png'))
+    >>> lena_hsv = color.rgb2hsv(lena)
     """
 
     if type(rgb) != np.ndarray:
@@ -25,10 +47,10 @@ def rgb2hsv(rgb):
     if rgb.ndim != 3 or rgb.shape[2] != 3:
         msg = "the input array 'rgb' must be have a shape == (.,.,3))"
         raise ValueError, msg
-              
+
     arr = rgb.astype("float32")
     out = np.empty_like(arr)
-    
+
     # -- V channel
     out_v = arr.max(-1)
 
@@ -39,15 +61,15 @@ def rgb2hsv(rgb):
 
     # -- H channel
     # red is max
-    idx = (arr[:,:,0] == out_v) 
+    idx = (arr[:,:,0] == out_v)
     out[idx, 0] = (arr[idx, 1] - arr[idx, 2]) / delta[idx]
 
     # green is max
-    idx = (arr[:,:,1] == out_v) 
+    idx = (arr[:,:,1] == out_v)
     out[idx, 0] = 2. + (arr[idx, 2] - arr[idx, 0] ) / delta[idx]
 
     # blue is max
-    idx = (arr[:,:,2] == out_v) 
+    idx = (arr[:,:,2] == out_v)
     out[idx, 0] = 4. + (arr[idx, 0] - arr[idx, 1] ) / delta[idx]
     out_h = (out[:,:,0] / 6.) % 1.