def scharr(data, sliceId=2):
edges = np.zeros(data.shape)
if sliceId == 2:
for idx in range(data.shape[2]):
edges[:, :, idx] = skifil.scharr(data[:, :, idx])
elif sliceId == 0:
for idx in range(data.shape[0]):
edges[idx, :, :] = skifil.scharr(data[idx, :, :])
return edges
def test_scharr_vertical():
"""Scharr on a vertical edge should be a vertical line."""
i, j = np.mgrid[-5:6, -5:6]
image = (j >= 0).astype(float)
result = filters.scharr(image) * np.sqrt(2)
j[np.abs(i) == 5] = 10000
assert_allclose(result[j == 0], 1)
assert (np.all(result[np.abs(j) > 1] == 0))
def test_scharr_horizontal():
"""Scharr on an edge should be a horizontal line."""
i, j = np.mgrid[-5:6, -5:6]
image = (i >= 0).astype(float)
result = filters.scharr(image) * np.sqrt(2)
# Fudge the eroded points
i[np.abs(j) == 5] = 10000
assert_allclose(result[i == 0], 1)
assert (np.all(result[np.abs(i) > 1] == 0))
def test_scharr_horizontal():
"""Scharr on an edge should be a horizontal line."""
i, j = np.mgrid[-5:6, -5:6]
image = (i >= 0).astype(float)
result = filters.scharr(image) * np.sqrt(2)
# Fudge the eroded points
i[np.abs(j) == 5] = 10000
assert_allclose(result[i == 0], 1)
assert (np.all(result[np.abs(i) > 1] == 0))
def test_scharr_horizontal():
"""Scharr on an edge should be a horizontal line."""
i, j = np.mgrid[-5:6, -5:6]
image = (i >= 0).astype(float)
result = filters.scharr(image) * np.sqrt(2)
# Check if result match transform direction
i[np.abs(j) == 5] = 10000
assert_allclose(result[i == 0], 1)
assert (np.all(result[np.abs(i) > 1] == 0))
def edge_method(method_num, im, sigma=2):
print("reached")
return {
1:feature.canny(im),
2:feature.canny(im, sigma=2),
3:roberts(im),
4:sobel(im),
5:scharr(im),
6:prewitt(im)
}[method_num]
def computeModels(img):
'''Calcule les resultats des differentes methodes de detection d'objets'''
#results = {'Image original': ing}
results = {}
results['Canny'] = feature.canny(img, sigma=1)
results['Roberts'] = roberts(img)
results['Sobel'] = sobel(img)
results['Scharr'] = scharr(img)
results['Prewitt'] = prewitt(img)
results['Laplace'] = laplace(img)
return results
def calculate_scharr(self, X, axis=0):
axis = axis
trans = []
for n in X:
p = []
for i in range(0, n.shape[axis]):
layer = n[:, :, i]
p.append(scharr(layer))
trans.append(np.asarray(p))
self.transformed = np.asarray(trans)
return self
def contour_map(self, image):
ratio = 0.75
image = gaussian(image, sigma=1)
contour = scharr(image)
contour = (contour - contour.min()) / (contour.max() - contour.min() +
1e-11)
mid_v = np.sort(contour.reshape(-1))
value = mid_v[int(ratio * len(mid_v))]
contour[contour < value] = 0
contour[contour >= value] = 1
return contour
def Get_Parameters(self):
copy_img = self.img
if len(self.img.shape) == 3:
copy_img = rgb2gray(self.img)
self.Variance = variance(laplace(copy_img, ksize=3))
self.MaxVariance = np.amax(laplace(copy_img, ksize=3))
self.Noise = estimate_sigma(copy_img)
self.Scharr = variance(scharr(copy_img))
print(self.Variance, self.MaxVariance, self.Scharr, self.Noise)
def mia_cmpedge(binary, verbose):
gradmag = scharr(binary)
thresh_gradmag = threshold_otsu(gradmag)
gradmag = gradmag > thresh_gradmag
gradmag = gradmag.astype(int)
gradmag_thin = skeletonize(gradmag)
plt.imshow(gradmag_thin)
gradmag_thin = gradmag_thin.astype(int)
[x, y] = np.nonzero(gradmag_thin)
[dx, dy] = smoothGradient(binary, 2)
return gradmag_thin, x, y, dx, dy
def read_mask(mask_id, shape):
mask_path = TRAIN_MASK_PATTERN.format(mask_id, mask_id)
masks = imread_collection(mask_path).concatenate()
height, width, _ = shape
num_masks = masks.shape[0]
mask = np.zeros((height, width), np.uint32)
dmask = np.zeros((height, width), np.uint32)
for index in range(0, num_masks):
dm = scharr(masks[index])
mask[masks[index] > 0] = 1
dmask[dm > 0] = 1
return np.dstack([mask, dmask])
def process_images(files):
n = len(files)
for i in range(n):
# Load image and use scharr to get edges
image = io.imread(files[i][0],as_grey=True)
image = scharr(img)
# Scale the intensity to whiten the edges
p95 = np.percentile(img, (95))
image = exposure.rescale_intensity(image, in_range=(0, p95))
return image
def binary_BF(image,
meanse=disk(10),
edgefilt='prewitt',
opense=disk(10),
fill_first=False,
bi_thresh=0.000025,
tophatse=disk(20)):
#convertim = img_as_ubyte(image)
meanim = rank.mean(image, meanse)
if edgefilt is 'prewitt':
edgeim = prewitt(meanim)
elif edgefilt is 'sobel':
edgeim = sobel(meanim)
elif edgefilt is 'scharr':
edgeim = scharr(meanim)
elif edgefilt is 'roberts':
edgeim = roberts(meanim)
closeim = closing(edgeim, opense)
openim = opening(closeim, opense)
if fill_first:
seed = np.copy(openim)
seed[1:-1, 1:-1] = openim.max()
mask = openim
filledim = reconstruction(seed, mask, method='erosion')
binarim = filledim > bi_thresh
else:
binarim = openim > bi_thresh * np.mean(openim)
seed = np.copy(binarim)
seed[1:-1, 1:-1] = binarim.max()
mask = binarim
filledim = reconstruction(seed, mask, method='erosion')
tophim = filledim - closing(white_tophat(filledim, tophatse),
opense) > 0.01
fig, ax = plt.subplots(nrows=2, ncols=4, figsize=(16, 8))
ax[0][0].imshow(image, cmap='gray')
ax[0][1].imshow(meanim, cmap='gray')
ax[0][2].imshow(edgeim, cmap='gray', vmax=4 * np.mean(edgeim))
ax[0][3].imshow(closeim, cmap='gray', vmax=4 * np.mean(closeim))
ax[1][0].imshow(openim, cmap='gray', vmax=4 * np.mean(openim))
ax[1][1].imshow(binarim, cmap='gray')
ax[1][2].imshow(filledim, cmap='gray')
ax[1][3].imshow(tophim, cmap='gray')
for axes in ax:
for axe in axes:
axe.axis('off')
fig.tight_layout()
return tophim
def execute(input_image, settings):
filter_number = settings.boundaries_settings[0]
grayscale = input_image.convert('L')
if filter_number == 1:
edges = roberts(grayscale)
elif filter_number == 2:
edges = prewitt(grayscale)
elif filter_number == 3:
edges = scharr(grayscale)
elif filter_number == 4:
edges = sobel(grayscale)
array = np.uint8(plt.cm.gist_earth(edges) * 255)
return Image.fromarray(array).convert(mode='L')
def get_contour_interior(mask):
if 'camunet' == config['param']['model']:
# 2-pixel contour (1out+1in), 2-pixel shrinked interior
outer = dilation(mask)
inner = erosion(mask)
contour = ((outer != inner) > 0).astype(np.uint8) * 255
interior = (erosion(inner) > 0).astype(np.uint8) * 255
else:
contour = filters.scharr(mask)
scharr_threshold = np.amax(abs(contour)) / 2.
contour = (np.abs(contour) > scharr_threshold).astype(np.uint8) * 255
interior = (mask - contour > 0).astype(np.uint8) * 255
return contour, interior
def addEdges(df, gray):
canny_edges = canny(gray, 0.6)
roberts_edges = roberts(gray)
sobel_edges = sobel(gray)
scharr_edges = scharr(gray)
prewitt_edges = prewitt(gray)
df['canny_edges'] = canny_edges.reshape(-1)
df['roberts_edge'] = roberts_edges.reshape(-1)
df['sobel_edge'] = sobel_edges.reshape(-1)
df['scharr_edge'] = scharr_edges.reshape(-1)
df['prewitt_edge'] = prewitt_edges.reshape(-1)
return df
def edge_entropy(img, bal=0.1):
"""
Weights pixels based on a weighted edge-detection and entropy balance.
Parameters
----------
img : ndarray
Image to weight.
bal : float (optional)
How much to value entropy (bal) versus edge-detection (1 - bal)
Returns
-------
ndarray :
Noramlized weight matrix for pixel sampling.
"""
dn_img = skimage.restoration.denoise_tv_bregman(img, 0.1)
img_gray = rgb2gray(dn_img)
img_lab = rgb2lab(dn_img)
entropy_img = gaussian(
img_as_float64(
dilation(entropy(img_as_ubyte(img_gray), disk(5)), disk(5))))
edges_img = dilation(
np.mean(np.array(
[scharr(img_lab[:, :, channel]) for channel in range(3)]),
axis=0), disk(3))
weight = (bal * entropy_img) + ((1 - bal) * edges_img)
weight /= np.mean(weight)
weight /= np.amax(weight)
if args.debug:
fig, (ax1, ax2, ax3) = plt.subplots(nrows=1,
ncols=3,
figsize=(8, 3),
sharex=True,
sharey=True)
ax1.imshow(entropy_img)
ax1.axis('off')
ax2.imshow(edges_img)
ax2.axis('off')
ax3.imshow(weight)
ax3.axis('off')
fig.tight_layout()
plt.show()
return weight
def loco(D, algorithm='muddy', detail=12, cutoff=0.04):
"""
Estimate local contrast for the given bitmap. Zero indicates a total
absence of local intensity variation, and one indicates ultimate contrast.
With other words, this sounds like a job for a traditional edge detector.
Parameters
----------
D : greyscale image array
Desaturated difference with the background.
algorithm : str, optional
The method to use. Choose between muddy (home grown, see notes below)
or one of Scharr / Sobel / Prewitt / Roberts (classics from scikits).
cutoff : float, optional
Eualization clipping limit, somewhere in the range 0 - 1.
Note that a value of 0.1 is already quite agressive.
detail : int or float, optional
A measure for the evaluation radius, in pixels.
Set this to the typical line width or edge gradient width.
Only relevant for the muddy algorithm.
Notes
-----
This is supposed to yield a map that approximates the intensity difference
between nearby light and dark zones for all points. Points with high local
contrast are eligible for serving as morph nodes. Our way of working here:
1. Normalize contrast through adaptive histogram equalization
2. Apply the selected algorithm. In case of muddy:
a) Subtract a Gaussian blur
b) Take the absolute value of this difference.
"""
G = desaturate(D, cutoff=cutoff)
algorithm = algorithm.lower().strip()
if algorithm == 'muddy':
F = gaussian_filter(G, sigma=detail)
C = abs(G - F)
elif algorithm == 'scharr':
C = scharr(G)
elif algorithm == 'sobel':
C = sobel(G)
elif algorithm == 'prewitt':
C = prewitt(G)
elif algorithm == 'roberts':
C = roberts(G)
return C
def preprocess():
images = []
for filename in os.listdir("/home/pi/self-driving/captured/all"):
if any([filename.endswith('.jpg')]):
img = imageio.imread(
os.path.join("/home/pi/self-driving/captured/all", filename))
if img is not None:
images.append(img)
for img in images:
img = rgb2gray(img)
img = img[200:480, 0:640]
edge_scharr = scharr(img)
imageio.imwrite("/home/pi/self-driving/processed/all/image.jpg",
edge_scharr)
def create_boundary_map(np_masks):
adjusted_masks = []
for i in range(np.shape(np_masks)[0]):
mask = np.squeeze(np_masks[i, :, :])
adjusted_mask = scharr(mask)
adjusted_mask[adjusted_mask > 0] = 1
adjusted_mask = adjusted_mask.astype(np.uint8)
kernel = np.ones((3, 3), np.uint8)
adjusted_mask = cv2.dilate(adjusted_mask, kernel, iterations=1)
adjusted_mask = np.invert(adjusted_mask) # * 255
adjusted_masks.append(adjusted_mask)
adjusted_masks = np.array(adjusted_masks)
return adjusted_masks - 254
def get_edges(self, detector='sobel'):
if detector == 'sobel':
img = filters.sobel(self.img_gray)
elif detector == 'canny1':
img = feature.canny(self.img_gray, sigma=1)
elif detector == 'canny3':
img = feature.canny(self.img_gray, sigma=3)
elif detector == 'scharr':
img = filters.scharr(self.img_gray)
elif detector == 'prewitt':
img = filters.prewitt(self.img_gray)
elif detector == 'roberts':
img = filters.roberts(self.img_gray)
return img
def match_ppl_xpl(data, crop_size, work_dir):
ppl_reference = scharr(data['ppl']['crops'][data['ppl']['ref-index']])
xpl_ref, xpl_ref_bbox, xpl_ref_shift = get_registered_averaged(
data['xpl'], crop_size)
delta_pair = [xpl_ref, crop_to_bounding_box(ppl_reference, xpl_ref_bbox)]
delta_xp = xpl_ref_shift + compute_displacements_direct(delta_pair, 0)[1]
ppl_reg_file = os.path.join(work_dir, "ref_dir_ppl.png")
cvt.cv2_to_file(cvt.simple_grayscale_stretch(ppl_reference), ppl_reg_file)
xpl_reg_file = os.path.join(work_dir, "ref_avg_xpl.png")
cvt.cv2_to_file(xpl_ref, xpl_reg_file)
return delta_xp
def get_contour_interior(mask, bold=False):
if True: #'camunet' == config['param']['model']:
# 2-pixel contour (1out+1in), 2-pixel shrinked interior
outer = binary_dilation(mask) #, square(9))
if bold:
outer = binary_dilation(outer) #, square(9))
inner = binary_erosion(mask) #, square(9))
contour = ((outer != inner) > 0).astype(np.uint8)
interior = (erosion(inner) > 0).astype(np.uint8)
else:
contour = filters.scharr(mask)
scharr_threshold = np.amax(abs(contour)) / 2.
contour = (np.abs(contour) > scharr_threshold).astype(np.uint8)
interior = (mask - contour > 0).astype(np.uint8)
return contour, interior
def build_gradient(image_data, band_list, gauss_sigma):
with warnings.catch_warnings():
warnings.simplefilter("ignore")
smooth_im_blue = ndimage.filters.gaussian_filter(
image_data[band_list[2]], sigma=gauss_sigma)
grad_image = img_as_ubyte(filters.scharr(smooth_im_blue))
# Prevent the watersheds from 'leaking' along the sides of the image
grad_image[:, 0] = grad_image[:, 1]
grad_image[:, -1] = grad_image[:, -2]
grad_image[0, :] = grad_image[1, :]
grad_image[-1, :] = grad_image[-2, :]
return grad_image
def contour_map(self, image):
ratio = 0.75
image = gaussian(image, sigma=1)
contour = scharr(image)
contour = (contour - contour.min()) / (contour.max() - contour.min() +
1e-11)
mid_v = np.sort(contour.reshape(-1))
value = mid_v[int(ratio * len(mid_v))]
# value = 0.1
# print(value)
contour[contour < value] = 0
contour[contour >= value] = 1
# cont = contour * 255
# cont = Image.fromarray(np.array(cont,np.uint8)).convert('L')
# cont.save(os.path.join("/root/chujiajia/Results/test/",str(contour.mean())+"img.pgm"))
return contour
def scharr_filter(image):
new_img = filters.scharr(image)
image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
new_img = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
fig, ax = plt.subplots(nrows=2)
ax[0].imshow(cv2.cvtColor(new_img, cv2.COLOR_BGR2RGB))
ax[0].set_title("Original image")
ax[1].imshow(cv2.cvtColor(new_img, cv2.COLOR_BGR2RGB))
ax[1].set_title("Image with Scharr filter")
plt.tight_layout()
plt.show()
def filter_show(data):
for k in [12, 3088]:
img = np.array(data[k]).reshape((28, 28))
image_scharr = scharr(img)
image_sobel = sobel(img)
image_prewitt = prewitt(img)
image_gabor_real, image_gabor_im = gabor(img, frequency=0.65)
image_roberts = roberts(img)
image_roberts_pos = roberts_pos_diag(img)
image_roberts_neg = roberts_neg_diag(img)
image_frangi = frangi(img)
image_laplace = laplace(img)
image_hessian = hessian(img)
image_threshold_local_3 = threshold_local(img, 3)
image_threshold_local_5 = threshold_local(img, 5)
image_threshold_local_7 = threshold_local(img, 7)
image_threshold_niblack = threshold_niblack(img, window_size=5, k=0.1)
image_threshold_sauvola = threshold_sauvola(img)
image_threshold_triangle = img > threshold_triangle(img)
fig, axes = plt.subplots(nrows=2,
ncols=2,
sharex=True,
sharey=True,
figsize=(8, 8))
ax = axes.ravel()
ax[0].imshow(img, cmap=plt.cm.gray)
ax[0].set_title('Original image')
ax[1].imshow(image_threshold_niblack, cmap=plt.cm.gray)
ax[1].set_title('Niblack')
ax[2].imshow(image_threshold_sauvola, cmap=plt.cm.gray)
ax[2].set_title('Sauvola')
ax[3].imshow(image_threshold_triangle, cmap=plt.cm.gray)
ax[3].set_title('Triangle')
for a in ax:
a.axis('off')
plt.tight_layout()
plt.show()
plt.close()
def apply_filter(img, name="equalizer"):
"""
applying filter to the input image
:param img: input image
:param name: requested filter
:return: image after applying the filter
"""
if name == "equalizer":
return exposure.equalize_hist(img)
elif name == "adaptive_equalization":
return exposure.equalize_adapthist(img)
elif name == "sobel_op":
return scharr(img, )
elif name == "canny_feature":
return canny(rgb2gray(img), sigma=4)
else:
return exposure.equalize_hist(img)
def __getitem__(self, index):
img = np.array(Image.open(self.data_path[index]))
lab = np.array(
Image.open(self.data_path[index].replace("image", "label").replace(
"Slide", "GT")).convert("L"))
img, lab = data_geneartor(img, lab, patch_size=self.patch_size)
marker = erosion(lab, square(3))
contour = filters.scharr(lab)
contour = dilation(contour, square(1))
contour[contour > 0] = 1
img = img.transpose((2, 0, 1))
img = torch.from_numpy(img).float()
lab = torch.from_numpy(np.array([lab])).float()
marker = torch.from_numpy(np.array([marker])).float()
contour = torch.from_numpy(np.array([contour])).float()
return img, lab, marker, contour
def all_feature_extractor(imgpath):
"""
Applies 58 filters and
returns the dictionary containing
the features obatined after applying on those images
"""
image = cv2.imread(imgpath)
image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
# Extracting Gabor Features
feature_dict = gabor_feature_extractor(image)
feature_dict['Original'] = image
entropy_img = entropy(image, disk(1))
feature_dict['Entropy'] = entropy_img
gaussian3_img = nd.gaussian_filter(image, sigma=3)
feature_dict['Gaussian3'] = gaussian3_img
gaussian7_img = nd.gaussian_filter(image, sigma=7)
feature_dict['Gaussian7'] = gaussian7_img
sobel_img = sobel(image)
feature_dict['Sobel'] = sobel_img
canny_edge_img = cv2.Canny(image, 100, 200)
feature_dict['Canny'] = canny_edge_img
robert_edge_img = roberts(image)
feature_dict['Robert'] = robert_edge_img
scharr_edge = scharr(image)
feature_dict['Scharr'] = scharr_edge
prewitt_edge = prewitt(image)
feature_dict['Prewitt'] = prewitt_edge
median_img = nd.median_filter(image, size=3)
feature_dict['Median'] = median_img
variance_img = nd.generic_filter(image, np.var, size=3)
feature_dict['Variance'] = variance_img
return feature_dict
def test4():
# im = np.array(Image.open("pentagon.png").convert("L"))
im = np.array(Image.open("/Users/MichaelMason/Desktop/original-small.jpg").convert("L"))
n, m = im.shape
edges = scharr(im)
# edges = gaussian(scharr(im))
plt.imshow(edges)
ixs = list(bresenham(0, n//2, m-1, n//2))
line = [float(ix) for ix, (j, i) in enumerate(ixs) if edges[i, j] > 0.1]
# line = [edges[i,j] for ix, (j, i) in enumerate(ixs)]
cluster_ixs = kdecluster(line)
print(cluster_ixs)
for i in cluster_ixs:
plt.plot(*ixs[int(np.round(i))], marker='x', color='k')
# plt.plot(*zip(*bresenham(0, n//2, m, n//2)))
plt.show()
def get_regions(slice_or_arr, fill_holes=False, clear_borders=True, threshold='otsu'):
"""Get the skimage regions of a black & white image."""
if threshold == 'otsu':
thresmeth = filters.threshold_otsu
elif threshold == 'mean':
thresmeth = np.mean
edges = filters.scharr(slice_or_arr.astype(np.float))
edges = filters.gaussian(edges, sigma=1)
thres = thresmeth(edges)
bw = edges > thres
if clear_borders:
segmentation.clear_border(bw, buffer_size=int(max(bw.shape)/50), in_place=True)
if fill_holes:
bw = ndimage.binary_fill_holes(bw)
labeled_arr, num_roi = measure.label(bw, return_num=True)
regionprops = measure.regionprops(labeled_arr, edges)
return labeled_arr, regionprops, num_roi
def preprocess(img):
img = img.convert('L')
bw_img = imgToGrayAndResize(img, 512)
scharr_image = arrayThreshold(scharr(bw_img), 0.04) # ocistime obrazok od slabych pixlov, aby sa trebars dal lahsie spoznat horizont
#canny_image = feature.canny(np.array(bw_img))
hog = feature.hog(scharr_image, orientations=16, pixels_per_cell=(32, 32), cells_per_block=(1, 1), normalise = True, visualise=False)
image_blocks = vectorBlockshaped(np.array(bw_img), 128, 128)
blocks_color_histogram = []
counter = 0
for block in image_blocks:
hist, bin_edges = np.histogram(block, bins=range(0,256))
blocks_color_histogram.append(hist)
blocks_color_histogram = np.array(blocks_color_histogram).ravel()
img_face_rotation_data, guessed_rotation_by_faces = np.array(getFaceCountByRotation(img))
return hog, blocks_color_histogram, img_face_rotation_data, guessed_rotation_by_faces
indistinguishable.
.. [1] https://en.wikipedia.org/wiki/Sobel_operator#Alternative_operators
.. [2] B. Jaehne, H. Scharr, and S. Koerkel. Principles of filter design. In
Handbook of Computer Vision and Applications. Academic Press, 1999.
.. [3] https://en.wikipedia.org/wiki/Prewitt_operator
"""
x, y = np.ogrid[:100, :100]
# Rotation-invariant image with different spatial frequencies
img = np.exp(1j * np.hypot(x, y)**1.3 / 20.).real
edge_sobel = sobel(img)
edge_scharr = scharr(img)
edge_prewitt = prewitt(img)
diff_scharr_prewitt = edge_scharr - edge_prewitt
diff_scharr_sobel = edge_scharr - edge_sobel
max_diff = np.max(np.maximum(diff_scharr_prewitt, diff_scharr_sobel))
fig, ((ax0, ax1), (ax2, ax3)) = plt.subplots(nrows=2, ncols=2, sharex=True, sharey=True, subplot_kw={'adjustable':'box-forced'})
ax0.imshow(img, cmap=plt.cm.gray)
ax0.set_title('Original image')
ax0.axis('off')
ax1.imshow(edge_scharr, cmap=plt.cm.gray)
ax1.set_title('Scharr Edge Detection')
ax1.axis('off')
def clusterDetector(imgstr, filetype, clustercolor):
def diameter(num_px):
if filetype=='tiff':
R = 0.16415411
elif filetype=='jpeg':
R = 24.95816
Apx = (50/R)**2 # area of single pixel, about 4 um^2 / sq px
A = Apx*num_px
return ((A/pi)**(0.5))*2
buf = cStringIO.StringIO(imgstr)
if clustercolor=='dark':
img = 255-imread(buf, as_grey=True)
elif clustercolor=='light':
img = imread(buf, as_grey=True)
print "Image read by worker, starting image processing"
med_img = filters.median(img, disk(3))
scharr_edges = filters.scharr(med_img)
edges = scharr_edges > filters.threshold_otsu(scharr_edges)
foreground = med_img > 50
watershed_image = closing(closing(foreground, disk(2)) ^ (edges), disk(2))
distance = ndimage.distance_transform_edt(watershed_image)
filtered_distance = ndimage.gaussian_filter(distance, 3)
local_maxi = peak_local_max(filtered_distance, indices=False,
min_distance=20, exclude_border=False)
markers = dilation(ndimage.label(local_maxi)[0], disk(10))
labels = watershed(-distance, markers, mask=erosion(watershed_image, disk(2)))
j = 1
cleared_labels = zeros_like(labels)
for i in range(1, labels.max()+1):
cleared = clear_border(labels==i).astype(int)
if cleared.max()>0:
cleared_labels += cleared * j
j += 1
props = regionprops(cleared_labels)
offset = 0
output = []
for i, prop in enumerate(props):
box = prop.bbox
fig = Figure()
canvas = FigureCanvasAgg(fig)
ax = fig.add_subplot(111)
ax.imshow(img[box[0]-offset:box[2]+offset, box[1]-offset:box[3]+offset], interpolation='nearest', cmap='Greys')
ax.imshow(prop.image, cmap="Reds", alpha=0.3)
ax.axis('off')
buf = cStringIO.StringIO()
fig.savefig(buf,format='png')
data = buf.getvalue().encode("base64").strip()
diam = diameter(prop.image.sum())
output.append({'data':data,'diameter':diam})
print "job finished"
return output
def test_scharr_zeros():
"""Scharr on an array of all zeros."""
result = filters.scharr(np.zeros((10, 10)), np.ones((10, 10), bool))
assert (np.all(result < 1e-16))
t2 = 160 t3 = 200 t4 = 200 s_Q = ( 2*imgQ + dstQ + t4 ) / ( imgQ**2 + dstQ**2 + t4 ) s_I = ( 2*imgI + dstI + t3 ) / ( imgI**2 + dstI**2 + t3 ) pc1 = phasepack.phasecong(imgY, nscale = 4, norient = 4, minWaveLength = 6, mult = 2, sigmaOnf=0.55) pc2 = phasepack.phasecong(dstY, nscale = 4, norient = 4, minWaveLength = 6, mult = 2, sigmaOnf=0.55) pc1 = pc1[0] pc2 = pc2[0] s_PC = ( 2*pc1 + pc2 + t1 ) / ( pc1**2 + pc2**2 + t1 ) g1 = scharr( imgY ) g2 = scharr( dstY ) s_G = ( 2*g1 + g2 + t2 ) / ( g1**2 + g2**2 + t2 ) s_L = s_PC * s_G s_C = s_I * s_Q pcM = np.maximum(pc1,pc2) fsim = round( np.nansum( s_L * pcM) / np.nansum(pcM), 3) fsimc = round( np.nansum( s_L * s_C**0.3 * pcM) / np.nansum(pcM), 3) print 'FSIM: ' + str(fsim) print 'FSIMC: ' + str(fsimc)
edge_sobel = sobel(img_name_gs)
fig, (ax0, ax1) = plt.subplots(ncols=2, sharex=True, sharey=True, subplot_kw={'adjustable':'box-forced'})
ax0.imshow(edge_roberts, cmap=plt.cm.gray)
ax0.set_title('Roberts Edge Detection')
ax0.axis('off')
ax1.imshow(edge_sobel, cmap=plt.cm.gray)
ax1.set_title('Sobel Edge Detection')
ax1.axis('off')
plt.tight_layout()
edge_sobel = sobel(img_name_gs)
edge_scharr = scharr(img_name_gs)
edge_prewitt = prewitt(img_name_gs)
diff_scharr_prewitt = edge_scharr - edge_prewitt
diff_scharr_sobel = edge_scharr - edge_sobel
max_diff = np.max(np.maximum(diff_scharr_prewitt, diff_scharr_sobel))
fig, ((ax0, ax1), (ax2, ax3)) = plt.subplots(nrows=2, ncols=2, sharex=True, sharey=True, subplot_kw={'adjustable':'box-forced'})
ax0.imshow(img_name_gs, cmap=plt.cm.gray)
ax0.set_title('Original image')
ax0.axis('off')
ax1.imshow(edge_scharr, cmap=plt.cm.gray)
ax1.set_title('Scharr Edge Detection')
ax1.axis('off')
def getEdge(img,mode=0):
return {0: sobel(img),
1:scharr(img),
2:prewitt(img),
3:roberts(img)}.get(mode, sobel(img))
plt.imshow(gauss2) lap=laplace(gray0,ksize=100) plt.imshow(lap) pre=prewitt(gray0, mask=None) plt.imshow(pre) pre_v=prewitt_v(gray0, mask=None) plt.imshow(pre_v) from skimage import filters edges2 = filters.roberts(gray0) plt.imshow(edges2) plt.imshow(scharr(gray0)) plt.imshow(threshold_mean(gray0)) plt.imshow(wiener(gray0)) ####################################### plt.imshow(img) plt.imshow(gray0) plt.imshow(image) ### TREES plt.imshow(segmentation) ### CONTOURS plt.imshow(img_back, cmap = 'gray') ### STREET plt.imshow(gy) plt.imshow(angle) plt.imshow(binary_adaptive)
def process(filename):
#
# Conventions:
# a_i, b_i
# are variables in image space, units are pixels
# a_w, b_w
# are variables in world space, units are meters
#
print '\n========================================'
print ' File: ' + filename
print '========================================\n'
im = imread(filename)
im = rgb2gray(im)
im = (im * 255.).astype(np.uint8)
tag_mosaic = TagMosaic(0.0254)
detections = AprilTagDetector().detect(im)
print ' %d tags detected.' % len(detections)
#
# Sort detections by distance to center
#
c_i = np.array([im.shape[1], im.shape[0]]) / 2.
dist = lambda p_i: np.linalg.norm(p_i - c_i)
closer_to_center = lambda d1, d2: int(dist(d1.c) - dist(d2.c))
detections.sort(cmp=closer_to_center)
mosaic_pos = lambda det: tag_mosaic.get_position_meters(det.id)
det_i = np.array([ d.c for d in detections ])
det_w = np.array([ mosaic_pos(d) for d in detections ])
#
# To learn a weighted local homography, we find the weighting
# function parameters that minimize reprojection error across
# leave-one-out validation folds of the data. Since the
# homography is local at the center, we only use 9 nearest
# detections to the center
#
det_i9 = det_i[:9]
det_w9 = det_w[:9]
def local_homography_loocv_error(theta, args):
src, tgt = args
errs = [ local_homography_error(theta, src[t_ix], tgt[t_ix], src[v_ix], tgt[v_ix])
for t_ix, v_ix in LeaveOneOut(len(src)) ]
return np.mean(errs)
def learn_homography_i2w():
result = minimize( local_homography_loocv_error,
x0=[ 50, 1, 1e-3 ],
args=[ det_i9, det_w9 ],
method='Powell',
options={'ftol': 1e-3} )
print '\nHomography: i->w'
print '------------------'
print ' params:', result.x
print ' rmse: %.6f' % sqrt(result.fun)
print '\n Optimization detail:'
print ' ' + str(result).replace('\n', '\n ')
H = create_local_homography_object(*result.x)
for i, w in zip(det_i9, det_w9):
H.add_correspondence(i, w)
return H
def learn_homography_w2i():
result = minimize( local_homography_loocv_error,
x0=[ 0.0254, 1, 1e-3 ],
method='Powell',
args=[ det_w9, det_i9 ],
options={'ftol': 1e-3} )
print '\nHomography: w->i'
print '------------------'
print ' params:', result.x
print ' rmse: %.6f' % sqrt(result.fun)
print '\n Optimization detail:'
print ' ' + str(result).replace('\n', '\n ')
H = create_local_homography_object(*result.x)
for w, i in zip(det_w9, det_i9):
H.add_correspondence(w, i)
return H
#
# We assume that the distortion is zero at the center of
# the image and we are interesting in the word to image
# homography at the center of the image. However, we don't
# know the center of the image in world coordinates.
# So we follow a procedure as explained below:
#
# First, we learn a homography from image to world
# Next, we find where the image center `c_i` maps to in
# world coordinates (`c_w`). Finally, we find the local
# homography `LH0` from world to image at `c_w`
#
H_iw = learn_homography_i2w()
c_i = np.array([im.shape[1], im.shape[0]]) / 2.
c_w = H_iw.map(c_i)[:2]
H_wi = learn_homography_w2i()
LH0 = H_wi.get_homography_at(c_w)
print '\nHomography at center'
print '----------------------'
print ' c_w =', c_w
print ' c_i =', c_i
print 'LH0 * c_w =', H_wi.map(c_w)
#
# Obtain distortion estimate
# detected + undistortion = mapped
# (or) undistortion = mapped - detected
#
def homogeneous_coords(arr):
return np.hstack([ arr, np.ones((len(arr), 1)) ])
mapped_i = LH0.dot(homogeneous_coords(det_w).T).T
mapped_i = np.array([ p / p[2] for p in mapped_i ])
mapped_i = mapped_i[:,:2]
undistortion = mapped_i - det_i # image + undistortion = mapped
max_distortion = np.max([np.linalg.norm(u) for u in undistortion])
print '\nMaximum distortion is %.2f pixels' % max_distortion
#
# Fit non-parametric model to the observations
#
model = GPModel(det_i, undistortion)
print '\nGP Hyper-parameters'
print '---------------------'
print ' x: ', model._gp_x._covf.theta
print ' log-likelihood: %.4f' % model._gp_x.model_evidence()
print ' y: ', model._gp_y._covf.theta
print ' log-likelihood: %.4f' % model._gp_y.model_evidence()
print ''
print ' Optimization detail:'
print ' [ x ]'
print ' ' + str(model._gp_x.fit_result).replace('\n', '\n ')
print ' [ y ]'
print ' ' + str(model._gp_y.fit_result).replace('\n', '\n ')
#
# Visualization
#
from skimage.filters import scharr
from matplotlib import pyplot as plt
plt.style.use('ggplot')
plt.figure(figsize=(16,10))
#__1__
plt.subplot(221)
plt.title(filename.split('/')[-1])
plt.imshow(im, cmap='gray')
plt.plot(det_i[:,0], det_i[:,1], 'o')
for i, d in enumerate(detections):
plt.text(d.c[0], d.c[1], str(i),
fontsize=8, color='white', bbox=dict(facecolor='maroon', alpha=0.75))
plt.grid()
plt.axis('equal')
#__2__
plt.subplot(222)
plt.title('Non-linear Homography')
if False:
plt.imshow(1.-scharr(im), cmap='bone', interpolation='gaussian')
from collections import defaultdict
row_groups = defaultdict(list)
col_groups = defaultdict(list)
for d in detections:
row, col = tag_mosaic.get_row(d.id), tag_mosaic.get_col(d.id)
row_groups[row] += [ d.id ]
col_groups[col] += [ d.id ]
for k, v in row_groups.iteritems():
a = tag_mosaic.get_position_meters(np.min(v))
b = tag_mosaic.get_position_meters(np.max(v))
x_coords = np.linspace(a[0], b[0], 100)
points = np.array([ H_wi.map([x, a[1]]) for x in x_coords ])
plt.plot(points[:,0], points[:,1], '-',color='#CF4457', linewidth=2)
for k, v in col_groups.iteritems():
a = tag_mosaic.get_position_meters(np.min(v))
b = tag_mosaic.get_position_meters(np.max(v))
y_coords = np.linspace(a[1], b[1], 100)
points = np.array([ H_wi.map([a[0], y]) for y in y_coords ])
plt.plot(points[:,0], points[:,1], '-',color='#CF4457', linewidth=2)
plt.plot(det_i[:,0], det_i[:,1], 'kx')
plt.grid()
plt.axis('equal')
#__3__
plt.subplot(223)
plt.title('Qualitative Estimates')
for i, d in enumerate(detections):
plt.text(d.c[0], d.c[1], str(i), fontsize=8, color='#999999')
X, Y = det_i[:,0], det_i[:,1]
U, V = undistortion[:,0], undistortion[:,1]
plt.quiver(X, Y, U, -V, units='dots')
# plt.quiver(X, Y, U, -V, angles='xy', scale_units='xy', scale=1) # plot exact
plt.text( 0.5, 0, 'max observed distortion: %.2f px' % max_distortion, color='#CF4457', fontsize=10,
horizontalalignment='center', verticalalignment='bottom', transform=plt.gca().transAxes)
plt.gca().invert_yaxis()
plt.axis('equal')
#__4__
plt.subplot(224)
plt.title('Qualitative Undistortion')
H, W = im.shape
grid = np.array([[x, y] for y in xrange(0, H, 80) for x in xrange(0, W, 80)])
predicted = model.predict(grid)
X, Y = grid[:,0], grid[:,1]
U, V = predicted[:,0], predicted[:,1]
plt.quiver(X, Y, U, -V, units='dots')
#plt.quiver(X, Y, U, -V, angles='xy', scale_units='xy', scale=1)) # plot exact
plt.gca().invert_yaxis()
plt.axis('equal')
plt.tight_layout()
plt.gcf().patch.set_facecolor('#dddddd')
#plt.show()
plt.savefig(filename + '.svg', bbox_inches='tight')
def frangi_segmentation(image,
colors,
frangi_args,
threshold_args,
separate_objects=True,
contrast_kernel_size='skip',
color_args_1='skip',
color_args_2='skip',
color_args_3='skip',
neighborhood_args='skip',
morphology_args_1='skip',
morphology_args_2='skip',
hollow_args='skip',
fill_gaps_args='skip',
diameter_args='skip',
diameter_bins='skip',
image_name='image',
verbose=False):
"""
Possible approach to object detection using frangi filters. Selects colorbands for
analysis, runs frangi filter, thresholds to identify candidate objects, then removes
spurrious objects by color and morphology characteristics. See frangi_approach.ipynb.
Unless noted, the dictionaries are called by their respective functions in order.
Parameters
----------
image : ndarray
RGB image to analyze
colors : dict or str
Parameters for picking the colorspace. See `pyroots.band_selector`.
frangi_args : list of dict or dict
Parameters to pass to `skimage.filters.frangi`
threshold_args : list of dict or dict
Parameters to pass to `skimage.filters.threshold_adaptive`
contrast_kernel_size : int, str, or None
Kernel size for `skimage.exposure.equalize_adapthist`. If `int`, then gives the size of the kernel used
for adaptive contrast enhancement. If `None`, uses default (1/8 shortest image dimension). If `skip`,
then skips.
color_args_1 : dict
Parameters to pass to `pyroots.color_filter`.
color_args_2 : dict
Parameters to pass to `pyroots.color_filter`. Combines with color_args_1
in an 'and' statement.
color_args_3 : dict
Parameters to pass to `pyroots.color_filter`. Combines with color_args_1, 2
in an 'and' statement.
neighborhood_args : dict
Parameters to pass to 'pyroots.neighborhood_filter'.
morphology_args_1 : dict
Parameters to pass to `pyroots.morphology_filter`
morphology_args_2 : dict
Parameters to pass to `pyroots.morphology_filter`. Happens after fill_gaps_args in the algorithm.
hollow_args : dict
Parameters to pass to `pyroots.hollow_filter`
fill_gaps_args : dict
Paramaters to pass to `pyroots.fill_gaps`
diameter_bins : list
To pass to `pyroots.bin_by_diameter`
image_name : str
Identifier of image for summarizing
Returns
-------
A dictionary containing:
1. `"geometry"` summary `pandas.DataFrame`
2. `"objects"` binary image
3. `"length"` medial axis image
4. `"diameter"` medial axis image
"""
# Pull band from colorspace
working_image = band_selector(image, colors) # expects dictionary (lazy coding)
nbands = len(working_image)
if verbose is True:
print("Color bands selected")
## Count nubmer of dictionaries in threshold_args and frangi_args. Should equal number of bands. Convert to list if necessary
try:
len(threshold_args[0])
except:
threshold_args = [threshold_args]
if nbands != len(threshold_args):
raise ValueError(
"""Number of dictionaries in `threshold_args` doesn't
equal the number of bands in `colors['band']`!"""
)
pass
try:
len(frangi_args[0])
except:
frangi_args = [frangi_args]
if nbands != len(frangi_args):
raise ValueError(
"""Number of dictionaries in `frangi_args` doesn't
equal the number of bands in `colors['band']`!"""
)
pass
working_image = [img_as_float(i) for i in working_image]
# Contrast enhancement
try:
for i in range(nbands):
temp = exposure.equalize_adapthist(working_image[i],
kernel_size = contrast_kernel_size)
working_image[i] = img_as_float(temp)
if verbose:
print("Contrast enhanced")
except:
if contrast_kernel_size is not 'skip':
warn('Skipping contrast enhancement')
pass
# invert if necessary
for i in range(nbands):
if not colors['dark_on_light'][i]:
working_image[i] = 1 - working_image[i]
# Identify smoothing sigma for edges and frangi thresholding
# simultaneously detect edges (computationally cheaper than multiple frangi enhancements)
edges = [np.ones_like(working_image[0]) == 1] * nbands # all True
sigma_val = [0.125] * nbands # step is 0, 0.25, 0.5, 1, 2, 4, 8, 16
for i in range(nbands):
edge_val = 1
while edge_val > 0.1 and sigma_val[i] < 10:
sigma_val[i] = 2*sigma_val[i]
temp = filters.gaussian(working_image[i], sigma=sigma_val[i])
temp = filters.scharr(temp)
temp = temp > filters.threshold_otsu(temp)
edge_val = np.sum(temp) / np.sum(np.ones_like(temp))
edges_temp = temp.copy()
if sigma_val[i] == 0.25: # try without smoothing
temp = filters.scharr(working_image[i])
temp = temp > filters.threshold_otsu(temp)
edge_val = np.sum(temp) / np.sum(np.ones_like(temp))
if edge_val <= 0.1:
sigma_val[i] = 0
edges_temp = temp.copy()
if separate_objects:
edges[i] = morphology.skeletonize(edges_temp)
if verbose:
print("Sigma value: {}".format(sigma_val))
if separate_objects:
print("Edges found")
# Frangi vessel enhancement
for i in range(nbands):
temp = filters.gaussian(working_image[i], sigma=sigma_val[i])
temp = filters.frangi(temp, **frangi_args[i])
temp = 1 - temp/np.max(temp)
temp = temp < filters.threshold_local(temp, **threshold_args[i])
working_image[i] = temp.copy()
frangi = working_image.copy()
if verbose:
print("Frangi filter, threshold complete")
# Combine bands, separate objects
combined = working_image[0] * ~edges[0]
for i in range(1, nbands):
combined = combined * working_image[i] * ~edges[i]
working_image = combined.copy()
# Filter candidate objects by color
try:
color1 = color_filter(image, working_image, **color_args_1) #colorspace, target_band, low, high, percent)
if verbose:
print("Color filter 1 complete")
except:
if color_args_1 is not 'skip':
warn("Skipping Color Filter 1")
color1 = np.ones(working_image.shape) # no filtering
try:
color2 = color_filter(image, working_image, **color_args_2) # nesting equates to an "and" statement.
if verbose:
print("Color filter 2 complete")
except:
if color_args_2 is not 'skip':
warn("Skipping Color Filter 2")
color2 = np.ones(working_image.shape) # no filtering
try:
color3 = color_filter(image, working_image, **color_args_3) # nesting equates to an "and" statement.
if verbose:
print("Color filter 3 complete")
except:
if color_args_3 is not 'skip':
warn("Skipping Color Filter 3")
color3 = np.ones(working_image.shape) # no filtering
# Combine bands
working_image = color1 * color2 * color3
del color1
del color2
del color3
# Re-expand to area
if separate_objects:
# find edges removed
temp = [frangi[i] * edges[i] for i in range(nbands)]
rm_edges = temp[0].copy()
for i in range(1, nbands):
rm_edges = rm_edges * temp[i]
# filter by color per criteria above
try: color1 = color_filter(image, rm_edges, **color_args_1)
except: color1 = np.ones(rm_edges.shape)
try: color2 = color_filter(image, rm_edges, **color_args_2)
except: color2 = np.ones(rm_edges.shape)
try: color3 = color_filter(image, rm_edges, **color_args_3)
except: color3 = np.ones(rm_edges.shape)
# Combine color filters
expanded = color1 * color2 * color3
else:
expanded = np.zeros(colorfilt.shape) == 1 # evaluate to false
working_image = expanded ^ working_image # bitwise or
try: # remove little objects (for computational efficiency)
working_image = morphology.remove_small_objects(
working_image,
min_size=morphology_args_1['min_size']
)
except:
pass
if verbose:
print("Edges re-added")
# Filter candidate objects by morphology
try:
working_image = morphology_filter(working_image, **morphology_args_1)
if verbose:
print("Morphology filter 1 complete")
except:
if morphology_args_1 is not 'skip':
warn("Skipping morphology filter 1")
pass
# Filter objects by neighborhood colors
try:
working_image = neighborhood_filter(image, working_image, **neighborhood_args)
if verbose:
print("Neighborhood filter complete")
except:
if neighborhood_args is not 'skip':
warn("Skipping neighborhood filter")
pass
# Filter candidate objects by hollowness
if hollow_args is not 'skip':
temp = morphology.remove_small_holes(working_image, min_size=10)
try:
if np.sum(temp) > 0:
working_image = hollow_filter(temp, **hollow_args)
if verbose:
print("Hollow filter complete")
except:
warn("Skipping hollow filter")
pass
# Close small gaps and holes in accepted objects
try:
working_image = fill_gaps(working_image, **fill_gaps_args)
if verbose:
print("Gap filling complete")
except:
if fill_gaps_args is not 'skip':
warn("Skipping filling gaps")
pass
# Filter candidate objects by morphology
try:
working_image = morphology_filter(working_image, **morphology_args_2)
if verbose:
print("Morphology filter 2 complete")
except:
if morphology_args_2 is not 'skip':
warn("Skipping morphology filter 2")
pass
# Skeletonize. Now working with a dictionary of objects.
skel = skeleton_with_distance(working_image)
if verbose:
print("Skeletonization complete")
# Diameter filter
try:
diam = diameter_filter(skel, **diameter_args)
if verbose:
print("Diameter filter complete")
except:
diam = skel.copy()
if diameter_args is not 'skip':
warn("Skipping diameter filter")
pass
# Summarize
if diameter_bins is None or diameter_bins is 'skip':
summary_df = summarize_geometry(diam['geometry'], image_name)
else:
diam_out, summary_df = bin_by_diameter(diam['length'],
diam['diameter'],
diameter_bins,
image_name)
diam['diameter'] = diam_out
out = {'geometry' : summary_df,
'objects' : diam['objects'],
'length' : diam['length'],
'diameter' : diam['diameter']}
if verbose is True:
print("Done")
return(out)
def ScharrMetric(self, field, unused_opts): # Scharr uses [3, 10, 3; 0, 0, 0; -3, -10, -3] return filters.scharr(numpy.abs(field))
def register_bands(image, template_band=1, ECC_criterion=True):
"""
Fix chromatic abberation in images by calculating and applying an affine
transformation. Chromatic abberation is a result of uneven refraction of light
of different wavelengths. It shows up as systematic blue and red edges in
high-contrast areas.
This should be done before other processing to minimize artifacts.
Parameters
----------
image : ndarray
3- or 4-channel image, probably RGB.
template : int
Band to which to align the other bands. Usually G in RGB.
ECC_criterion : bool
Use ECC criterion to find optimal warp? Improves results, but increases
processing time 5x.
Returns
-------
An ndarray of `image.size`
Notes
-----
Uses `skimage.filters.scharr` to find edges in each band, then finds and
applies an affine transformation to register the images using
`cv2.estimateRigidTransform` and `cv2.warpAffine`. If `ECC_criterion=True`,
the matrix from `estimateRigidTransform` is updated using `cv2.findTransformECC`.
"""
#find dimensions
height, width, depth = image.shape
#define bands to analyze
analyze = []
for i in range(depth):
if i != template_band:
analyze.append(i)
# Extract bands, find edges
bands = img_split(image)
edges = [np.ones_like(i) for i in bands]
for i in range(len(bands)):
sigma_val = 0.25
edge_val = 1
while edge_val > 0.1 and sigma_val < 10:
temp = filters.gaussian(bands[i], sigma=sigma_val)
scharr = filters.scharr(temp)
temp = scharr > filters.threshold_otsu(scharr)
edge_val = np.sum(temp) / np.sum(np.ones_like(temp))
sigma_val = 2*sigma_val
edges[i] = img_as_ubyte(scharr * temp)
#make output image
out = np.zeros((height, width, depth), dtype=np.uint8)
out[:, :, template_band] = bands[template_band]
try:
for i in analyze:
# Estimate transformation
warp_matrix = np.array(cv2.estimateRigidTransform(edges[template_band],
edges[i],
fullAffine=False), dtype=np.float32)
if ECC_criterion == True:
# Optimize using ECC criterion and default settings
warp_matrix = cv2.findTransformECC(edges[template_band],
edges[i],
warpMatrix=warp_matrix)[1]
# transform
aligned = cv2.warpAffine(bands[i],
warp_matrix,
(width, height),
flags=cv2.INTER_LINEAR + cv2.WARP_INVERSE_MAP, # otherwise the transformation goes the wrong way
borderMode=cv2.BORDER_CONSTANT)
# add to color image
out[:, :, i] = aligned
except:
# Probably few objects, so no smoothing and no thresholding to have as much info as possible
edges = [img_as_ubyte(filters.scharr(i)) for i in edges]
for i in analyze:
# Estimate transformation
warp_matrix = np.array(cv2.estimateRigidTransform(edges[template_band],
edges[i],
fullAffine=False), dtype=np.float32)
if ECC_criterion == True:
# Optimize using ECC criterion and default settings
warp_matrix = cv2.findTransformECC(edges[template_band],
edges[i],
warpMatrix=warp_matrix)[1]
# transform
aligned = cv2.warpAffine(bands[i],
warp_matrix,
(width, height),
flags=cv2.INTER_LINEAR + cv2.WARP_INVERSE_MAP, # otherwise the transformation goes the wrong way
borderMode=cv2.BORDER_CONSTANT)
# add to color image
out[:, :, i] = aligned
return(img_as_ubyte(out))
def _scharr(self, img):
# Invert the image to ease edge detection.
img = 1. - img
grey = skc.rgb2grey(img)
return skf.scharr(grey)
def _compute_gradient_image(self):
from skimage import filters
self.edges = filters.scharr(self._image)
def test_scharr_mask():
"""Scharr on a masked array should be zero."""
np.random.seed(0)
result = filters.scharr(np.random.uniform(size=(10, 10)),
np.zeros((10, 10), bool))
assert_allclose(result, 0)
