medial_axis function results with some double lines and small holes

[yxdragon]

Description

[I use the medial_axis function, the result image has some double lines or small holes, then I count make a topology analysis!]
![demo picture] (http://data.imagepy.org/skebug.png "The medial_axis result and mine result")

I view the source code, and found the medial_axis function, sort by distance map, then use the skeletonize function's table to check if to remove the pixcel. But some times the sort cannot make sure the out-inside turn in extreme condition. So we need a new table for the media_axis. This is my code：

Build the table

from scipy.ndimage import label, generate_binary_structure

strc = generate_binary_structure(2, 2)

# check whether this pixcel can be removed
def check(n):
    a = [(n>>i) & 1 for i in range(8)]
    a.insert(4, 0) 
    ''' make the 3x3 unit
       0  1  2
       3  p  4
       5  6  7
    '''
    # if up, down, left, right all are 1, you cannot make a hole
    if a[1] & a[3] & a[5] & a[7]:return False
    a = np.array(a).reshape((3,3))
    # segments
    n = label(a, strc)[1]
    # if sum is 0, it is a isolate point, you cannot remove it.
    # if sum is 1, it is a terminal point, you cannot remove it.
    # if number of segments = 2, you cannot split the line.
    # if number of segments > 2, it is a node, you cannot remove it.
    return a.sum()>1 and n<2

lut = [check(n) for n in range(256)]
'''
like this:
array([[0, 0, 0, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 0, 1, 1],
       [0, 0, 1, 1, 1, 0, 1, 1, 0, 0, 1, 1, 0, 0, 1, 1],
       [0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 1, 1],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 1],
       [0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 1, 1],
       [1, 0, 1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 1, 1, 0, 0],
       [1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 1, 1],
       [1, 0, 1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 1, 1, 0, 0],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
       [1, 0, 1, 1, 1, 0, 1, 1, 0, 0, 1, 1, 0, 0, 1, 1],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 1],
       [1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 1, 1],
       [1, 0, 1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 1, 1, 0, 0],
       [1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 1, 1],
       [1, 0, 1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 1, 1, 0, 0]])
'''

Remove the pixcels

def skel2d(data, idx, lut):
    h, w = data.shape
    data = data.flat
    for i in idx:
        if data[i]==0 : continue
        xy = [i-w-1, i-w, i-w+1,\
                i-1  ,            i+1  ,\
                i+w-1, i+w, i+w+1]

        c = sum([(data[xy[j]]==255)<<j for j in range(8)])
        if lut[c]:data[i] = 128

def mid_axis(img):
    dis = nimg.distance_transform_edt(img)
    idx = np.argsort(dis.flat).astype(np.int32)
    skel2d(img, idx, lut)

I think my code is right, but needs more test.
May I send a Pull Request after rewrite in cython?

Way to reproduce

[If reporting a bug, please include the following important information:]

• [test image ] (http://data.imagepy.org/ske_bug.png "My Test Image")
• [Ubuntu 14.0 ] Operating system and version
• [ 2.7] Python version
• [ 0.12.3] scikit-image version (run skimage.__version__)

[stefanv]

@yxdragon Sorry for not noticing this issue sooner; can you give us a bit more information here? I'm trying to understand what is wrong, and how you propose we fix it.