def main():
"""Load image, apply filter, plot."""
img = data.camera()
# just like sobel, but no -2/+2 in the middle
prewitt_y = np.array([
[-1, -1, -1],
[0, 0, 0],
[1, 1, 1]
])
prewitt_x = np.rot90(prewitt_y) # rotates counter-clockwise
img_sx = signal.correlate(img, prewitt_x, mode="same")
img_sy = signal.correlate(img, prewitt_y, mode="same")
g_magnitude = np.sqrt(img_sx**2 + img_sy**2)
ground_truth = skifilters.prewitt(data.camera())
util.plot_images_grayscale(
[img, img_sx, img_sy, g_magnitude, ground_truth],
["Image", "Prewitt (x)", "Prewitt (y)", "Prewitt (both/magnitude)",
"Prewitt (Ground Truth)"]
)
def process_frame(frame, ref, wettedArea, theta, threshold1, threshold2):
"""
Compare the frame of interest to a reference frame and already known
wetted area and determine what additional areas have becomed wetted using
the absolute difference and the difference in edge detection.
"""
# For testing and optimizing, use these flags to turn off either step
simpleSubtraction = True
edgeSubtraction = True
cutoff = 254
# Initialize storage variables
image1 = np.zeros_like(frame)
image2 = np.zeros_like(frame)
tempRef = np.zeros_like(frame)
comp1 = np.zeros_like(ref)
comp2 = np.zeros_like(ref)
# Generate comparison data between reference and image
if simpleSubtraction:
image1 = subtract_images(frame, ref)
if edgeSubtraction:
tempFrame = np.uint8(filters.prewitt(frame) * 255)
tempRef = np.uint8(filters.prewitt(ref) * 255)
image2 = subtract_images(tempFrame, tempRef)
# Prepare matrices for thresholding the results
# Apply different thresholds at different intensities
comp1[:] = threshold1
comp2[:] = threshold2
# comp1[ref>=30] = threshold1
# comp1[ref<30] = 2*threshold1
# comp2[tempRef<=128] = threshold2
# comp2[tempRef>128] = 2*threshold2
# comp2[tempRef<30] = threshold2*.75
# Convert the results to 8 bit images for combining
image1 = np.uint8((image1 > comp1) * 255)
image2 = np.uint8((image2 > threshold2) * 255)
# wettedArea = np.uint8(wettedArea*255)
# Depending on whether or not the disk is rotating, apply thresholding
if theta != 0:
image1 = rotate_image(image1, theta, size=image1.shape)
image2 = rotate_image(image2, theta, size=image2.shape)
wettedArea = wettedArea + (image1 > cutoff) + (image2 > cutoff)
# Fill holes in the wetted area
# wettedArea = ndimage.morphology.binary_fill_holes(wettedArea)
return wettedArea
def highpass_filter(image, what):
newImage = np.array(image)
if newImage.ndim == 3:
Filters = {
'roberts': [
filters.roberts(image[:, :, 0]),
filters.roberts(image[:, :, 1]),
filters.roberts(image[:, :, 2])
],
'sobel': [
filters.sobel(image[:, :, 0]),
filters.sobel(image[:, :, 1]),
filters.sobel(image[:, :, 2])
],
'scharr': [
filters.scharr(image[:, :, 0]),
filters.scharr(image[:, :, 1]),
filters.scharr(image[:, :, 2])
],
'prewitt': [
filters.prewitt(image[:, :, 0]),
filters.prewitt(image[:, :, 1]),
filters.prewitt(image[:, :, 2])
],
'laplace': [
filters.laplace(image[:, :, 0]),
filters.laplace(image[:, :, 1]),
filters.laplace(image[:, :, 2])
],
'canny': [
feature.canny(image[:, :, 0]),
feature.canny(image[:, :, 1]),
feature.canny(image[:, :, 2])
]
}
newImageR = Filters[what][0]
newImageG = Filters[what][1]
newImageB = Filters[what][2]
newImage = np.array(newImageR + newImageG + newImageB)
else:
Filters = {
'roberts': filters.roberts(image),
'sobel': filters.sobel(image),
'scharr': filters.scharr(image),
'prewitt': filters.prewitt(image),
'laplace': filters.laplace(image),
'canny': feature.canny(image)
}
newImage = Filters[what]
return np.clip(newImage, 0, 255)
def _gms(x_pred, x_real, c):
"""
:rtype: np.ndarray
"""
def rgb2gray(rgb):
r, g, b = rgb[:, :, 0], rgb[:, :, 1], rgb[:, :, 2]
gray = 0.2989 * r + 0.5870 * g + 0.1140 * b
return gray
grad_pred = np.zeros(shape=x_pred.shape[:3])
grad_real = np.zeros(shape=x_real.shape[:3])
if len(x_pred.shape) == 4 and x_pred.shape[-1] == 3:
for i in range(x_pred.shape[0]):
grad_pred[i, :, :] = prewitt(rgb2gray(x_pred[i, :, :, :]))
grad_real[i, :, :] = prewitt(rgb2gray(x_real[i, :, :, :]))
elif len(x_pred.shape) == 3 and x_pred.shape[-1] == 3:
grad_pred = prewitt(rgb2gray(x_pred))
grad_real = prewitt(rgb2gray(x_real))
elif len(x_pred.shape) == 4 and x_pred.shape[-1] == 1:
for i in range(x_pred.shape[0]):
grad_pred[i, :, :] = prewitt(x_pred[i, :, :, 0])
grad_real[i, :, :] = prewitt(x_real[i, :, :, 0])
elif len(x_pred.shape) == 3 and x_pred.shape[-1] == 1:
grad_pred = prewitt(x_pred)
grad_real = prewitt(x_real)
else:
raise ValueError(
'Check the function in the source code. You might need to develop this condition.'
)
gms = (2 * grad_real * grad_pred + c) / (grad_real**2 + grad_pred**2 + c)
return gms
def getEdge(img, mode=0):
return {
0: sobel(img),
1: scharr(img),
2: prewitt(img),
3: roberts(img)
}.get(mode, sobel(img))
def show_diff_edge_detectors(img):
matplotlib.rcParams['font.size'] = 8
edge_roberts = roberts(img)
edge_sobel = sobel(img)
edge_scharr = scharr(img)
edge_prewitt = prewitt(img)
canny = feature.canny(img, sigma=0.5)
fig, axes = plt.subplots(nrows=2, ncols=3, sharex=True, sharey=True,
figsize=(10, 7))
ax = axes.ravel()
ax[0].imshow(img)
ax[0].set_title('Input image')
ax[1].imshow(edge_prewitt)
ax[1].set_title('Prewitt Edge Detection')
ax[2].imshow(edge_scharr)
ax[2].set_title('Scharr Edge Detection')
ax[3].imshow(edge_sobel)
ax[3].set_title('Sobel Edge Detection')
ax[4].imshow(edge_roberts)
ax[4].set_title('Roberts Edge Detection')
ax[5].imshow(canny)
ax[5].set_title('Canny Edge Detection')
for a in ax:
a.axis('off')
plt.tight_layout()
plt.show()
def search_satellite(self, point, polygon):
polygon_path = Path(polygon, closed=True) # ?
image = self.utils.center_image("satellite", point,
self.zoom) # /samples/staticmap2
masked_image = self.utils.mask_image(image, polygon) # masked image
grayscale_image = self.utils.grayscale(
masked_image) # grayscaled masked image
lot_paths = []
edge_image = filters.prewitt(
grayscale_image) # image with prewitt operator applied
edge_regions = self.regions.search(edge_image) # 4 region boxes
for region in edge_regions:
region_image = masked_image[region[0][0]:region[1][0],
region[0][1]:region[1]
[1], :] # splice image into regions
profile = self.profiler.profile(region_image) # ???
if profile == PolygonProfile.LOT:
bbox = self.utils.region_to_bbox(region)
lot_paths.append(polygon_path.clip_to_bbox(bbox, inside=True))
if self.debug and len(lot_paths) > 0:
self.utils.draw_path_on_image(image, lot_paths)
return lot_paths
def filter():
phasenames = ['train', 'val']
tr = 0
idx = 0
cnttxt = 0
cntnon = 0
for phase in [0, 1]:
f = open('%s_unbalance.txt' % (phasenames[phase]), 'r')
f2 = open('%s.txt' % (phasenames[phase]), 'w')
for s in f:
if s.split(' ')[1][0] == '0':
imgname = '%s%s' % (base, s.split(' ')[0])
print(imgname)
im = io.imread(imgname, as_grey=True)
e = prewitt(im)
emax = np.amax(e) / 2
et = np.reshape(
np.array([0 if e1 < emax else 1 for e1 in e.flatten()]),
(32, 32))
ecnt = np.sum(et)
if ecnt > thresh:
f2.write(s)
tr = tr + (1 if (phase == 0) else 0)
idx = idx + 1
cntnon = cntnon + 1
else:
f2.write(s)
tr = tr + (1 if (phase == 0) else 0)
idx = idx + 1
cnttxt = cnttxt + 1
f.close()
f2.close()
print('total=', idx, ' training=', tr, ' Val=', idx - tr)
print('Text=', cnttxt, ' non-Text=', cntnon)
def test_prewitt_vertical():
"""Prewitt on a vertical edge should be a vertical line."""
i, j = np.mgrid[-5:6, -5:6]
image = (j >= 0).astype(float)
result = filters.prewitt(image) * np.sqrt(2)
assert_allclose(result[j == 0], 1)
assert_allclose(result[np.abs(j) > 1], 0, atol=1e-10)
def test_prewitt_mask(dtype):
"""Prewitt on a masked array should be zero."""
result = filters.prewitt(
np.random.uniform(size=(10, 10)).astype(dtype),
np.zeros((10, 10), dtype=bool))
assert result.dtype == _supported_float_type(dtype)
assert_allclose(np.abs(result), 0)
def preprocess_and_extract_features(data):
'''Preprocess data and extract features
Preprocess: normalize, scale, repair
Extract features: transformations and dimensionality reduction
'''
# Here, we do something trivially simple: we take the average of the RGB
# values to produce a grey image, transform that into a vector, then
# extract the mean and standard deviation as features.
# Make the image grayscale
#data = np.mean(data, axis=3)
data1 = []
for row in range(data.shape[0]):
data_4loop = prewitt(data[row])
data1.append(data_4loop)
data = np.array(data1)
data[data > 140] = 255
#np.clip(data, 0, 170, out=data)
# Vectorize the grayscale matrices
vectorized_data = data.reshape(data.shape[0], -1)
# extract the mean and standard deviation of each sample as features
feature_mean = np.mean(vectorized_data, axis=1)
feature_std = np.std(vectorized_data, axis=1)
feature_variance = np.var(vectorized_data, axis=1)
# RED Combine the extracted features into a single feature vector
features = np.stack((feature_mean, feature_std), axis=-1)
return features
def border_thing(Array, n):
# Array ya está recortado
res = []
for joder in range(n):
img = Array[joder]
img_g = rgb2gray(img)
# aplicación de filtros
for k in range(0, 10):
img_g = gaussian(img_g, 3)
img_prw = prewitt(img_g, mask=None)
#figure();
#imshow(img_prw, cmap = "gray");
#title("prewitt");
#figure();
#hist(img_prw.ravel(), 256, [0,256]);
us_img = (255 * img_prw < 0.67).astype("uint8")
#figure();
#imshow(us_img, cmap = "gray");
#title("umbralización");
# conteo
count = 0
fil, col, can = img.shape
for c in range(0, fil):
for j in range(0, col):
if (us_img[c][j] == 0):
count += 1
res.append(count)
return res
def skeleton_image(file):
image_origin = io.imread(file, as_grey=True)
image_origin = np.where(image_origin > 0, 1, 0)
image_origin = adjunction_image(image_origin)
# 骨架
edge_roberts = roberts(image_origin)
edge_roberts = feature.canny(image_origin, sigma = 1)
edge_roberts = np.where(edge_roberts > 0, 3, 0)
skeleton = skeletonize_3d(image_origin)
# back_image = np.where(image_origin > 0, 0, 1)
back_image = np.where(image_origin > 0, 0, 1)
back_thin_images = skeletonize(back_image)
[row, col] = back_thin_images.shape
back_thin_image = np.zeros((row, col))
back_thin_image[:, int(col / 5):int(col / 5) * 4] = back_thin_images[:, int(col / 5):int(col / 5) * 4]
p2 = plt.subplot(421)
p2.imshow(image_origin + back_thin_images, cmap="gray")
total_image = edge_roberts + back_thin_images
p2 = plt.subplot(422)
p2.imshow(total_image)
p3 = plt.subplot(425)
p3.imshow(np.where(roberts(image_origin) > 0, 1, 0))
p3 = plt.subplot(426)
p3.imshow(np.where(sobel(image_origin) > 0, 2, 0))
p3 = plt.subplot(427)
p3.imshow(np.where(scharr(image_origin) > 0, 2, 0))
p3 = plt.subplot(428)
p3.imshow(np.where(prewitt(image_origin) > 0, 2, 0))
plt.show()
def calc_potential(img, params):
c = 5
# img_g = skifil.gaussian_filter(img, params['sigma'])
img_g = skires.denoise_tv_chambolle(img, weight=params['tv_weight'])
# img_g = skifil.denoise_bilateral(img, sigma_range=params['sigma_range'], sigma_spatial=params['sigma_spatial'])
grad_1 = skifil.prewitt(img_g)
grad_2 = skifil.roberts(img_g)
grad_3 = skifil.scharr(img_g)
grad_4 = skifil.sobel(img_g)
# plt.figure()
# hist, bins = skiexp.histogram(img_g)
# plt.plot(bins, hist)
# plt.show()
pot = c * np.abs(grad_1)
plt.figure()
max_v = 0.0005
plt.subplot(221), plt.imshow(grad_1, 'gray',
interpolation='nearest'), plt.title('grad')
plt.subplot(222), plt.imshow(grad_2,
'gray',
interpolation='nearest',
vmax=max_v), plt.title('grad, adjusted vmax')
plt.subplot(224), plt.imshow(
pot, 'gray', interpolation='nearest'), plt.title('potential')
# plt.show()
return img_g
def edges(im):
original = skimage.io.imread(im)
grayscale = rgb2gray(original)
edge_sobel = filters.sobel(grayscale)
edge_scharr = filters.scharr(grayscale)
edge_prewitt = filters.prewitt(grayscale)
diff_scharr_prewitt = compare_images(edge_scharr, edge_prewitt)
diff_scharr_sobel = compare_images(edge_scharr, edge_sobel)
max_diff = np.max(np.maximum(diff_scharr_prewitt, diff_scharr_sobel))
fig, axes = plt.subplots(nrows=2,
ncols=2,
sharex=True,
sharey=True,
figsize=(8, 8))
axes = axes.ravel()
# axes[0].imshow(im, cmap=plt.cm.gray)
# axes[0].set_title('Original image')
axes[1].imshow(edge_scharr, cmap=plt.cm.gray)
axes[1].set_title('Scharr Edge Detection')
axes[2].imshow(diff_scharr_prewitt, cmap=plt.cm.gray, vmax=max_diff)
axes[2].set_title('Scharr - Prewitt')
axes[3].imshow(diff_scharr_sobel, cmap=plt.cm.gray, vmax=max_diff)
axes[3].set_title('Scharr - Sobel')
for ax in axes:
ax.axis('off')
plt.tight_layout()
plt.show()
def test_prewitt_horizontal():
"""Prewitt on an edge should be a horizontal line."""
i, j = np.mgrid[-5:6, -5:6]
image = (i >= 0).astype(float)
result = filters.prewitt(image) * np.sqrt(2)
# Check if result match transform direction
assert (np.all(result[i == 0] == 1))
assert_allclose(result[np.abs(i) > 1], 0, atol=1e-10)
def test_prewitt_vertical():
"""Prewitt on a vertical edge should be a vertical line."""
i, j = np.mgrid[-5:6, -5:6]
image = (j >= 0).astype(float)
result = filters.prewitt(image) * np.sqrt(2)
j[np.abs(i) == 5] = 10000
assert_allclose(result[j == 0], 1)
assert_allclose(result[np.abs(j) > 1], 0, atol=1e-10)
def sharp1(img, enhance, x):
y = x.get()
if (y == 1):
img = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
lap = cv2.Laplacian(img, cv2.CV_64F, ksize=5)
im = Image.fromarray(lap)
imgtk = ImageTk.PhotoImage(image=im, master=enhance)
imglabel = tk.Label(enhance, image=imgtk)
imglabel.image = imgtk
imglabel.place(x=100, y=150)
elif (y == 2):
img = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
n = img_as_float(img)
robert = roberts(n)
noise = img_as_ubyte(robert)
im = Image.fromarray(noise)
imgtk = ImageTk.PhotoImage(image=im, master=enhance)
imglabel = tk.Label(enhance, image=imgtk)
imglabel.image = imgtk
imglabel.place(x=100, y=150)
elif (y == 3):
img = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
n = img_as_float(img)
sobe = sobel(n)
noise = img_as_ubyte(sobe)
im = Image.fromarray(noise)
imgtk = ImageTk.PhotoImage(image=im, master=enhance)
imglabel = tk.Label(enhance, image=imgtk)
imglabel.image = imgtk
imglabel.place(x=100, y=150)
elif (y == 4):
img = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
edges = cv2.Canny(img, 100, 200)
im = Image.fromarray(edges)
imgtk = ImageTk.PhotoImage(image=im, master=enhance)
imglabel = tk.Label(enhance, image=imgtk)
imglabel.image = imgtk
imglabel.place(x=100, y=150)
elif (y == 5):
img = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
n = img_as_float(img)
pre = prewitt(n)
noise = img_as_ubyte(pre)
im = Image.fromarray(noise)
imgtk = ImageTk.PhotoImage(image=im, master=enhance)
imglabel = tk.Label(enhance, image=imgtk)
imglabel.image = imgtk
imglabel.place(x=100, y=150)
elif (y == 6):
n = img_as_float(img)
ma = unsharp_mask(n, radius=2, amount=2)
noise = img_as_ubyte(ma)
im = Image.fromarray(noise)
imgtk = ImageTk.PhotoImage(image=im, master=enhance)
imglabel = tk.Label(enhance, image=imgtk)
imglabel.image = imgtk
imglabel.place(x=100, y=150)
def sys_prewitt(img):
'''
prewitt系统自带
:param img: 原始图像
:return: 返回边缘图像
'''
img = cv2.imread(img, 0)
edge_img = filters.prewitt(img)
return edge_img
def test_prewitt_horizontal():
"""Prewitt on an edge should be a horizontal line."""
i, j = np.mgrid[-5:6, -5:6]
image = (i >= 0).astype(float)
result = filters.prewitt(image) * np.sqrt(2)
# Fudge the eroded points
i[np.abs(j) == 5] = 10000
assert (np.all(result[i == 0] == 1))
assert_allclose(result[np.abs(i) > 1], 0, atol=1e-10)
def test_prewitt_horizontal():
"""Prewitt on an edge should be a horizontal line."""
i, j = np.mgrid[-5:6, -5:6]
image = (i >= 0).astype(float)
result = filters.prewitt(image) * np.sqrt(2)
# Fudge the eroded points
i[np.abs(j) == 5] = 10000
assert (np.all(result[i == 0] == 1))
assert_allclose(result[np.abs(i) > 1], 0, atol=1e-10)
def show_img_edge():
img = data.camera()
edge_prewitt = prewitt(img)
edge_sobel = sobel(img)
fig, (ax1, ax2) = plt.subplots(ncols=2, figsize=(10, 5))
# Prewitt 边缘检测
ax1.imshow(edge_prewitt, cmap=plt.cm.gray)
ax2.imshow(edge_sobel, cmap=plt.cm.gray)
plt.show()
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 calculate_prewitt(self, X):
trans = []
for n in X:
p = []
for i in range(0, 50, 3):
layer = n[:, :, i]
p.append(prewitt(layer))
trans.append(np.asarray(p))
self.transformed = np.asarray(trans)
return self
def prewitt_rgb(image):
""" Aplica o filtro Prewitt, utilizando o decorador
@adapt_rgb para que o filtro seja aplicado em todos
os canais RGB da imagem.
@param image numpy.ndarray.
@return result numpy.ndarray.
"""
return filters.prewitt(image)
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 Filtro4(matrix_imagem):
#prewit
#matrix_imagem=Filtro11(matrix_imagem)
mask = np.array([[0, 1, 0], [1, 0, 1], [0, 1, 0]])
mat_empty = np.zeros_like(matrix_imagem - matrix_imagem.min(),
dtype="uint16")
for i in xrange(matrix_imagem.shape[0]):
mat_empty[i] = filters.prewitt(matrix_imagem[i].astype("uint16") -
matrix_imagem.min())
return mat_empty.astype('int16') + matrix_imagem.min()
def get_result(self):
'''
The resulting image for each project (Picture 4 on the page).
'''
arr = np.array(self.grayscale_pil)
if self.pid == 1:
if self.choice == 'ostu':
self.option_value = threshold_otsu(arr)
elif self.choice == 'entropy':
self.option_value = threshold_li(arr)
else:
pass
result_arr = (arr > self.option_value) * 255
elif self.pid == 2:
if self.choice == 'ED:Rob':
result_arr = (1 - roberts(arr)) * 255.
elif self.choice == 'ED:Prew':
result_arr = (1 - prewitt(arr)) * 255.
elif self.choice == 'ED:Sob':
result_arr = (1 - sobel(arr)) * 255.
elif self.choice == 'NR:gaus':
result_arr = gaussian(arr, sigma=self.option_value) * 255.
elif self.choice == 'NR:med':
result_arr = median(arr, disk(self.option_value))
else:
pass
elif self.pid == 3:
fn_dict = {
"dilation": dilation,
"erosion": erosion,
"opening": opening,
"closing": closing
}
fn = fn_dict[self.choice]
result_arr = fn(arr, disk(self.option_value))
elif self.pid == 4:
thresh = threshold_otsu(arr)
binary_arr = arr > thresh
if self.choice == "distance/ostu":
skel, distance = medial_axis(binary_arr, return_distance=True)
max_distance = np.max(distance)
min_distance = np.min(distance)
result_arr = (distance - min_distance) / (max_distance -
min_distance) * 255.
if self.choice == 'skeleton/ostu':
skel = skeletonize(binary_arr)
result_arr = skel * 255.
else:
result_arr = arr
result_pil = PIL.Image.fromarray((result_arr).astype(np.uint8))
return serve_pil(result_pil)
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 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 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 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
io.show() ## from skimage.filters import gaussian, gaussian_filter, laplace,prewitt from skimage.filters import prewitt_v,prewitt_h,scharr, wiener gauss=gaussian(gray0, sigma=5, multichannel=True) plt.imshow(gauss) gauss2=gaussian_filter(gray0, sigma=5, multichannel=True) 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)
.. [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 prewitt(X):
pw = filters.prewitt(X)
return pw
import nifty.cgp as ncgp
import nifty.segmentation as nseg
import skimage.data as sdata
import skimage.filters as sfilt
# plotting
import pylab
# get data
img = sdata.coins()
# edge indicator
edgeIndicator = sfilt.prewitt(sfilt.gaussian(img, 3))
# watersheds
overseg = nseg.seededWatersheds(edgeIndicator,
method='node_weighted', acc='min')
f = pylab.figure()
f.add_subplot(2,2,1)
pylab.imshow(img/255)
f.add_subplot(2,2,2)
pylab.imshow(edgeIndicator)
f.add_subplot(2,2,3)
pylab.imshow(overseg, cmap=nseg.randomColormap(overseg.max()+1))
f.add_subplot(2,2,4)
pylab.imshow(nseg.markBoundaries(img, overseg))
pylab.show()
def test_prewitt_mask():
"""Prewitt on a masked array should be zero."""
np.random.seed(0)
result = filters.prewitt(np.random.uniform(size=(10, 10)),
np.zeros((10, 10), bool))
assert_allclose(np.abs(result), 0)
def getEdge(img,mode=0):
return {0: sobel(img),
1:scharr(img),
2:prewitt(img),
3:roberts(img)}.get(mode, sobel(img))
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 test_prewitt_zeros():
"""Prewitt on an array of all zeros."""
result = filters.prewitt(np.zeros((10, 10)), np.ones((10, 10), bool))
assert_allclose(result, 0)
def detect_edges(imagename, algorithm):
'''
Does edge detection by the choosen algorithm.
:param imagename: image name
:param algorithm: has to be "roberts", "scharr", "prewitt", "canny-1", "canny-2" or "canny3"
:returns: image
'''
im = color.rgb2gray(imagename) # image is colored, lets make it gray scale
logging.info(algorithm + ' was choosen as edge detection algorithm.')
if algorithm == "roberts":
edges = filters.roberts(im)
elif algorithm == "scharr":
edges = filters.scharr(im)
elif algorithm == "sobel":
edges = filters.sobel(im)
elif algorithm == "prewitt":
edges = filters.prewitt(im)
elif algorithm == "canny-1":
edges = feature.canny(im, sigma=1)
elif algorithm == "canny-2":
edges = feature.canny(im, sigma=2)
elif algorithm == "canny-3":
edges = feature.canny(im, sigma=3)
return edges