Python morphology.skeletonize函数代码示例

def edge_detect(depth, color):
    # Get the gradient direction from the depth image
    graddir = grad_dir(depth)
    # plt.imshow(graddir)
    # plt.show()

    # kernel for dilation
    kernel = np.ones((5, 5), np.uint8)

    # Threshold the image so it is in the RGB color space
    bw2 = (((graddir - graddir.min()) / (graddir.max() - graddir.min())) * 255.9).astype(np.uint8)

    # removes the salt and pepper noise
    # by replacing pixels with
    # the median value of the area
    median = cv2.medianBlur(bw2, 9)

    # find edges with the canny edge detector
    bw2 = auto_canny(median)
    dilation2 = cv2.dilate(bw2, kernel, iterations=1)
    skel2 = morphology.skeletonize(dilation2 > 0)

    # Now run canny edge detector on the colour image
    # create a CLAHE object (Arguments are optional).
    # this does adaptive histogram equalization on the image
    clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8, 8))
    cl1 = clahe.apply(color)
    # median = cv2.medianBlur(bw2,5)
    # bw1 = cv2.GaussianBlur(cl1, (3,3), 0)

    # Perform canny edge detection on colour image, twice
    # 1. Detect outlines and fill with close function
    # 2. Detect outlines of now filled contoured image
    bw1 = auto_canny(cl1)
    closing = cv2.morphologyEx(bw1, cv2.MORPH_CLOSE, kernel, iterations=6)
    # dilation1 = cv2.dilate(bw1,kernel,iterations = 1)
    # skel1 = morphology.skeletonize(dilation1 > 0)
    bw1 = auto_canny(closing)

    # combine the edges from the color image and the depth image
    orop = (np.logical_or(bw1, skel2)).astype('uint8')

    # display results
    plt.subplot(1, 2, 1), plt.imshow(graddir, cmap='jet')
    plt.title('gradient dir'), plt.xticks([]), plt.yticks([])
    plt.subplot(1, 2, 2), plt.imshow(median, cmap='gray')
    plt.title('blurred image'), plt.xticks([]), plt.yticks([])
    plt.show()

    # dilate and skeletonize on combined image
    kernel = np.ones((5, 5), np.uint8)
    dilation = cv2.dilate(orop, kernel, iterations=1)
    img_out = morphology.skeletonize(dilation > 0)

    return img_out

def get_mask(img_org):
    noise_ratio = 0.3
    img = img_org > noise_ratio * img_org.max()
    skeleton = skeletonize(img)
    
    mask = dilation(skeleton > 0, disk(2))
    return mask

 def get_descriptor_lvX(self,img):
     ori = img
     #img = cv2.bitwise_not(numpy.array(img))
     #img = threshold_adaptive(numpy.array(img), 40)
     #img = cv2.bitwise_not(img*255.)
     img = skeletonize(numpy.array(img)/255.)*255.
     '''figure()
     gray()
     subplot(221)
     imshow(ori)
     subplot(222)
     imshow(img)
     show()'''
     #e = stats.entropy(img.flatten())
     #if math.isnan(e) or math.isinf(e):
     #    return 0
     #else:
     #    return e
     descs = hog(numpy.array(img), orientations=4, pixels_per_cell=(10, 10),cells_per_block=(3, 3),visualise=False)
     '''figure()
     gray()
     imshow(img)
     figure()
     imshow(hpgimg)
     show()'''
     return descs

def polylinesFromBinImage(img, minimum_cluster_size=6,
                          remove_small_obj_size=3,
                          reconnect_size=3,
                          max_n_contours=None, max_len_contour=None,
                          copy=True):
    '''
    return a list of arrays of un-branching contours

    img -> (boolean) array 

    optional:
    ---------
    minimum_cluster_size -> minimum number of pixels connected together to build a contour

    ##search_kernel_size -> TODO
    ##min_search_kernel_moment -> TODO

    numeric:
    -------------
    max_n_contours -> maximum number of possible contours in img
    max_len_contour -> maximum contour length

    '''
    assert minimum_cluster_size > 1
    assert reconnect_size % 2, 'ksize needs to be odd'

    # assert search_kernel_size == 0 or search_kernel_size > 2 and search_kernel_size%2, 'kernel size needs to be odd'
    # assume array size parameters, is not given:
    if max_n_contours is None:
        max_n_contours = max(img.shape)
    if max_len_contour is None:
        max_len_contour = sum(img.shape[:2])
    # array containing coord. of all contours:
    contours = np.zeros(shape=(max_n_contours, max_len_contour, 2),
                        dtype=np.uint16)  # if not search_kernel_size else np.float32)

    if img.dtype != np.bool:
        img = img.astype(bool)
    elif copy:
        img = img.copy()

    if remove_small_obj_size:
        remove_small_objects(img, remove_small_obj_size,
                             connectivity=2, in_place=True)
    if reconnect_size:
        # remove gaps
        maximum_filter(img, reconnect_size, output=img)
        # reduce contour width to 1
        img = skeletonize(img)

    n_contours = _populateContoursArray(img, contours, minimum_cluster_size)
    contours = contours[:n_contours]

    l = []
    for c in contours:
        ind = np.zeros(shape=len(c), dtype=bool)
        _getValidInd(c, ind)
        # remove all empty spaces:
        l.append(c[ind])
    return l

def bottleneck_distribution(image):
    """
    Count the distribution of bottlenecks

    :param image: data (binary)
    :type image: :py:class:`numpy.ndarray`

    :return: count of bottlenecks of size 4 and size 2
    :rtype: tuple(int)
    """
    skel = morphology.skeletonize(image)
    # get the distances
    dists = ndimage.distance_transform_edt(image)
    # ok for all the nonzero in the skeleton, we get the distances
    x_nz, y_nz = skel.nonzero() # get all the nonzero indices
    width4 = 0
    width2 = 0
    for i in range(len(x_nz)):
        x = x_nz[i]
        y = y_nz[i]

        dist = dists[x,y]
        if dist <= 4:
            width4 += 1
        if dist <= 2:
            width2 += 1

    return width4, width2

def skeletonize_mitochondria(mch_channel):
    mch_collector = np.max(mch_channel, axis=0)  # TODO: check max projection v.s. sum
    skeleton_labels = np.zeros(mch_collector.shape, dtype=np.uint8)

    # thresh = np.max(mch_collector)/2.
    thresh = threshold_otsu(mch_collector)
    # use adaptative threshold? => otsu seems to be sufficient in this case

    skeleton_labels[mch_collector > thresh] = 1
    skeleton2 = skeletonize(skeleton_labels)
    skeleton, distance = medial_axis(skeleton_labels, return_distance=True)
    active_threshold = np.mean(mch_collector[skeleton_labels]) * 5

    # print active_threshold
    transform_filter = np.zeros(mch_collector.shape, dtype=np.uint8)
    transform_filter[np.logical_and(skeleton > 0, mch_collector > active_threshold)] = 1
    skeleton = transform_filter * distance

    skeleton_ma = np.ma.masked_array(skeleton, skeleton > 0)
    skeleton_convolve = ndi.convolve(skeleton_ma, np.ones((3, 3)), mode='constant', cval=0.0)
    divider_convolve = ndi.convolve(transform_filter, np.ones((3, 3)), mode='constant', cval=0.0)
    skeleton_convolve[divider_convolve > 0] = skeleton_convolve[divider_convolve > 0] / \
                                              divider_convolve[divider_convolve > 0]
    new_skeleton = np.zeros_like(skeleton)
    new_skeleton[skeleton2] = skeleton_convolve[skeleton2]
    skeleton = new_skeleton

    return skeleton_labels, mch_collector, skeleton, transform_filter

def ruler_scale_factor(image, distance):
    """Returns the scale factor to convert from image coordinates to real world coordinates

    Args:
        image: BGR image of shape n x m x 3.
        distance: The real world size of the smallest graduation spacing
    Returns:
        float: Unitless scale factor from image coordinates to real world coordinates.

    """

    height, width = image.shape[:2]
    image, mask = find_ruler(image)
    binary_image = mask * threshold(image, mask)

    if binary_image[mask].mean() > 0.5:
        binary_image[mask] = ~binary_image[mask]
    remove_large_components(binary_image, max(height, width))
    edges = skeletonize(binary_image)
    hspace, angles, distances = hough_transform(edges)
    features = hspace_features(hspace, splits=16)
    angle_index = best_angles(np.array(features))

    max_graduation_size = int(max(image.shape))
    line_separation_pixels = find_grid(hspace[:, angle_index], max_graduation_size)

    logging.info('Line separation: {:.3f}'.format(line_separation_pixels))
    return distance / line_separation_pixels

def skeletonize(mask):
    """Reduces binary objects to 1 pixel wide representations (skeleton)

    Inputs:
    mask       = Binary image data

    Returns:
    skeleton   = skeleton image

    :param mask: numpy.ndarray
    :return skeleton: numpy.ndarray
    """
    # Store debug
    debug = params.debug
    params.debug = None

    # Convert mask to boolean image, rather than 0 and 255 for skimage to use it
    skeleton = skmorph.skeletonize(mask.astype(bool))

    skeleton = skeleton.astype(np.uint8) * 255

    # Reset debug mode
    params.debug = debug
    # Auto-increment device
    params.device += 1

    if params.debug == 'print':
        print_image(skeleton, os.path.join(params.debug_outdir, str(params.device) + '_skeleton.png'))
    elif params.debug == 'plot':
        plot_image(skeleton, cmap='gray')

    return skeleton

def skeletonize_mitochondria(mCh_channel):

    mch_collector = np.max(mCh_channel, axis=0)  # TODO: check how max affects v.s. sum
    labels = np.zeros(mch_collector.shape, dtype=np.uint8)

    # thresh = np.max(mch_collector)/2.
    thresh = threshold_otsu(mch_collector)
    # TODO: use adaptative threshold? => otsu seems to be sufficient in this case
    # http://scikit-image.org/docs/dev/auto_examples/xx_applications/plot_thresholding.html#sphx
    # -glr-auto-examples-xx-applications-plot-thresholding-py
    #  log-transform? => Nope, does not work
    # TODO: hessian/laplacian of gaussian blob detection?

    labels[mch_collector > thresh] = 1
    skeleton2 = skeletonize(labels)
    skeleton, distance = medial_axis(labels, return_distance=True)
    active_threshold = np.mean(mch_collector[labels]) * 5

    # print active_threshold
    transform_filter = np.zeros(mch_collector.shape, dtype=np.uint8)
    transform_filter[np.logical_and(skeleton > 0, mch_collector > active_threshold)] = 1
    skeleton = transform_filter * distance

    skeleton_ma = np.ma.masked_array(skeleton, skeleton > 0)
    skeleton_convolve = ndi.convolve(skeleton_ma, np.ones((3, 3)), mode='constant', cval=0.0)
    divider_convolve = ndi.convolve(transform_filter, np.ones((3, 3)), mode='constant', cval=0.0)
    skeleton_convolve[divider_convolve > 0] = skeleton_convolve[divider_convolve > 0] \
                                              / divider_convolve[divider_convolve > 0]
    new_skeleton = np.zeros_like(skeleton)
    new_skeleton[skeleton2] = skeleton_convolve[skeleton2]
    skeleton = new_skeleton

    return labels, mch_collector, skeleton, transform_filter

def get_thinned(binaryArr):
    """
    Return thinned output
    Parameters
    ----------
    binaryArr : Numpy array
        2D or 3D binary numpy array

    Returns
    -------
    result : boolean Numpy array
        2D or 3D binary thinned numpy array of the same shape
    """
    assert np.max(binaryArr) in [0, 1], "input must always be a binary array"
    voxCount = np.sum(binaryArr)
    if binaryArr.sum() == 0:
        return binaryArr.astype(bool)
    elif len(binaryArr.shape) == 2:
        return skeletonize(binaryArr).astype(bool)
    else:
        start_skeleton = time.time()
        zOrig, yOrig, xOrig = np.shape(binaryArr)
        orig = np.lib.pad(binaryArr, 1, 'constant')
        result = cy_get_thinned3D(np.uint64(orig))
        print("thinned %i number of pixels in %0.2f seconds" % (voxCount, time.time() - start_skeleton))
        return result[1:zOrig + 1, 1: yOrig + 1, 1: xOrig + 1].astype(bool)

 def calculate(self, image: np.ndarray, disk_size: int=9,
               mean_threshold: int=100, min_object_size: int=750) -> float:
     # Find edges that have a strong vertical direction
     vertical_edges = sobel_v(image)
     # Separate out the areas where there is a large amount of vertically-oriented stuff
     segmentation = self._segment_edge_areas(vertical_edges, disk_size, mean_threshold, min_object_size)
     # Draw a line that follows the center of the segments at each point, which should be roughly vertical
     # We should expect this to give us four approximately-vertical lines, possibly with many gaps in
     # each line
     skeletons = skeletonize(segmentation)
     # Use the Hough transform to get the closest lines that approximate those four lines
     hough = transform.hough_line(skeletons, np.arange(-constants.FIFTEEN_DEGREES_IN_RADIANS,
                                                       constants.FIFTEEN_DEGREES_IN_RADIANS,
                                                       0.0001))
     # Create a list of the angles (in radians) of all of the lines the Hough transform produced, with 0.0
     # being completely vertical
     # These angles correspond to the angles of the four sides of the channels, which we need to
     # correct for
     angles = [angle for _, angle, dist in zip(*transform.hough_line_peaks(*hough))]
     if not angles:
         raise ValueError("Image rotation could not be calculated. Check the images to see if they're weird.")
     else:
         # Get the average angle and convert it to degrees
         offset = sum(angles) / len(angles) * 180.0 / math.pi
         if offset > constants.ACCEPTABLE_SKEW_THRESHOLD:
             log.warn("Image is heavily skewed. Check that the images are valid.")
         return offset

def label_image(image):
    
    ROI = np.zeros((470,400,3), dtype=np.uint8)
    for c in range(3):
        for i in range(50,520):
            for j in range(240,640):
                ROI[i-50,j-240,c] = image[i,j,c]

    
    gray_ROI = cv2.cvtColor(ROI,cv2.COLOR_BGR2GRAY)
    
    ROI_flou = cv2.medianBlur((ROI).astype('uint8'),3)
    
    Laser = Detecte_laser.Detect_laser(ROI_flou)
    
    open_laser = cv2.morphologyEx(Laser, cv2.MORPH_DILATE, disk(3))
    
    skel = skeletonize(open_laser > 0)
    
    tranche = Detecte_laser.tranche(skel,90,30)    
    
    ret, thresh = cv2.threshold(gray_ROI*tranche.astype('uint8'),0,1,cv2.THRESH_BINARY_INV+cv2.THRESH_OTSU)
    thresh01 = thresh<1.0
    
    open_thresh = cv2.morphologyEx(thresh01.astype('uint8'), cv2.MORPH_OPEN, disk(10))
    
    labelised = (label(open_thresh,8,0))+1
    
    return gray_ROI,labelised

def label_particles_edge(im, sigma=2, closing_size=0, **extra_args):
    """ Segment image using Canny edge-finding filter.

        parameters
        ----------
        im : image in which to find particles
        sigma : size of the Canny filter
        closing_size : size of the closing filter

        returns
        -------
        labels : an image array of uniquely labeled segments
    """
    from skimage.morphology import square, binary_closing, skeletonize
    if skimage_version < StrictVersion('0.11'):
        from skimage.filter import canny
    else:
        from skimage.filters import canny
    edges = canny(im, sigma=sigma)
    if closing_size > 0:
        edges = binary_closing(edges, square(closing_size))
    edges = skeletonize(edges)
    labels = sklabel(edges)
    print "found {} segments".format(labels.max())
    # in ma.array mask, False is True, and vice versa
    labels = np.ma.array(labels, mask=edges == 0)
    return labels

    def enhance_edges(self):
        """ GNIRS_edge_detector method to enhance the footprint edges using the Sobel
            kernel. Generate two binary images, one with the left edges
            and the other showing the right edges only. This is because the
            MDF information about the footprints location is not well 
            determined.
        """
        sdata=nd.sobel(self.image,axis=self.axis)
        std = np.std(sdata)
        bdata=np.where(sdata>std,1,0)
 
        # Make the edges one pixel wide
        self.left_bin_image = skeletonize(bdata)

        bdata=np.where(sdata < -std,1,0)
        self.right_bin_image = skeletonize(bdata)

 def test_skeletonize_already_thinned(self):
     im = np.zeros((5, 5), np.uint8)
     im[3, 1:-1] = 1
     im[2, -1] = 1
     im[4, 0] = 1
     result = skeletonize(im)
     numpy.testing.assert_array_equal(result, im)