def extract_efd_features(image, n=10):
contours = get_contours(image)
if len(contours) == 0:
return np.zeros((4 * n))[3:]
efd = elliptic_fourier_descriptors(contours[0], order=n, normalize=True)
return efd.flatten()[3:]
def extract_efd_features(image, n=10):
contours = get_contours(image)
if len(contours) == 0:
return np.zeros((4*n))[3:]
efd = elliptic_fourier_descriptors(contours[0], order=n, normalize=True)
return efd.flatten()[3:]
def iter_blob_extremes(image, n=5):
original_shape = image.shape[::-1]
if max(original_shape) < 2000:
size = (500, 500)
y_scale = original_shape[0] / 500
x_scale = original_shape[1] / 500
else:
size = (1000, 1000)
y_scale = original_shape[0] / 1000
x_scale = original_shape[1] / 1000
img = resize(image, size)
bimg = gaussian_filter(img, sigma=1.0)
bimg = threshold_adaptive(bimg, 20, offset=2 / 255)
bimg = -bimg
bimg = ndi.binary_fill_holes(bimg)
label_image = label(bimg, background=False)
label_image += 1
regions = regionprops(label_image)
regions.sort(key=attrgetter('area'), reverse=True)
iter_n = 0
for region in regions:
try:
iter_n += 1
if iter_n > n:
break
# Skip small images
if region.area < int(np.prod(size) * 0.05):
continue
coords = get_contours(
add_border(label_image == region.label,
size=label_image.shape,
border_size=1,
background_value=False))[0]
coords = np.fliplr(coords)
top_left = sorted(coords,
key=lambda x: np.linalg.norm(np.array(x)))[0]
top_right = sorted(coords,
key=lambda x: np.linalg.norm(
np.array(x) - [img.shape[1], 0]))[0]
bottom_left = sorted(coords,
key=lambda x: np.linalg.norm(
np.array(x) - [0, img.shape[0]]))[0]
bottom_right = sorted(
coords,
key=lambda x: np.linalg.norm(
np.array(x) - [img.shape[1], img.shape[0]]))[0]
scaled_extremes = [(int(x[0] * y_scale), int(x[1] * x_scale))
for x in (top_left, top_right, bottom_left,
bottom_right)]
yield scaled_extremes
except Exception:
pass
raise SudokuExtractError("No suitable blob could be found.")
def iter_blob_extremes(image, n=5):
original_shape = image.shape[::-1]
if max(original_shape) < 2000:
size = (500, 500)
y_scale = original_shape[0] / 500
x_scale = original_shape[1] / 500
else:
size = (1000, 1000)
y_scale = original_shape[0] / 1000
x_scale = original_shape[1] / 1000
img = resize(image, size)
bimg = gaussian_filter(img, sigma=1.0)
bimg = threshold_adaptive(bimg, 20, offset=2/255)
bimg = -bimg
bimg = ndi.binary_fill_holes(bimg)
label_image = label(bimg, background=False)
label_image += 1
regions = regionprops(label_image)
regions.sort(key=attrgetter('area'), reverse=True)
iter_n = 0
for region in regions:
try:
iter_n += 1
if iter_n > n:
break
# Skip small images
if region.area < int(np.prod(size) * 0.05):
continue
coords = get_contours(add_border(label_image == region.label,
size=label_image.shape,
border_size=1,
background_value=False))[0]
coords = np.fliplr(coords)
top_left = sorted(coords, key=lambda x: np.linalg.norm(np.array(x)))[0]
top_right = sorted(coords, key=lambda x: np.linalg.norm(np.array(x) - [img.shape[1], 0]))[0]
bottom_left = sorted(coords, key=lambda x: np.linalg.norm(np.array(x) - [0, img.shape[0]]))[0]
bottom_right = sorted(coords, key=lambda x: np.linalg.norm(np.array(x) - [img.shape[1], img.shape[0]]))[0]
scaled_extremes = [(int(x[0] * y_scale), int(x[1]*x_scale)) for x in (top_left, top_right, bottom_left, bottom_right)]
yield scaled_extremes
except Exception:
pass
raise SudokuExtractError("No suitable blob could be found.")
def iter_blob_contours(image, n=5):
original_shape = image.shape[::-1]
if max(original_shape) < 2000:
size = (500, 500)
y_scale = original_shape[0] / 500
x_scale = original_shape[1] / 500
else:
size = (1000, 1000)
y_scale = original_shape[0] / 1000
x_scale = original_shape[1] / 1000
img = resize(image, size)
bimg = gaussian_filter(img, sigma=1.0)
bimg = threshold_adaptive(bimg, 20, offset=2 / 255)
bimg = (~binary_erosion(bimg))
label_image = label(bimg, background=False)
label_image += 1
regions = regionprops(label_image)
regions.sort(key=attrgetter('area'), reverse=True)
iter_n = 0
for region in regions:
iter_n += 1
if iter_n > n:
break
try:
coords = get_contours(
add_border(label_image == region.label,
size=label_image.shape,
border_size=1,
background_value=False))[0]
if np.linalg.norm(coords[0, :] - coords[-1, :]) > 1e-10:
raise SudokuExtractError("Not a closed contour.")
else:
coords = np.fliplr(coords[:-1, :])
top_left = sorted(coords,
key=lambda x: np.linalg.norm(np.array(x)))[0]
top_right = sorted(coords,
key=lambda x: np.linalg.norm(
np.array(x) - [img.shape[1], 0]))[0]
bottom_left = sorted(coords,
key=lambda x: np.linalg.norm(
np.array(x) - [0, img.shape[0]]))[0]
bottom_right = sorted(
coords,
key=lambda x: np.linalg.norm(
np.array(x) - [img.shape[1], img.shape[0]]))[0]
tl_i = np.argmax((coords == top_left).sum(axis=1))
tr_i = np.argmax((coords == top_right).sum(axis=1))
bl_i = np.argmax((coords == bottom_left).sum(axis=1))
br_i = np.argmax((coords == bottom_right).sum(axis=1))
coords[:, 0] *= y_scale
coords[:, 1] *= x_scale
if tl_i > bl_i:
left_edge = coords[bl_i:tl_i + 1, :]
else:
coords_end_of_array = coords[bl_i:, :]
coords_start_of_array = coords[:tl_i + 1]
left_edge = np.concatenate(
[coords_end_of_array, coords_start_of_array], axis=0)
if tr_i > tl_i:
top_edge = coords[tl_i:tr_i + 1, :]
else:
coords_end_of_array = coords[tl_i:, :]
coords_start_of_array = coords[:tr_i + 1]
top_edge = np.concatenate(
[coords_end_of_array, coords_start_of_array], axis=0)
if br_i > tr_i:
right_edge = coords[tr_i:br_i + 1, :]
else:
coords_end_of_array = coords[tr_i:, :]
coords_start_of_array = coords[:br_i + 1]
right_edge = np.concatenate(
[coords_end_of_array, coords_start_of_array], axis=0)
if bl_i > br_i:
bottom_edge = coords[br_i:bl_i + 1, :]
else:
coords_end_of_array = coords[br_i:, :]
coords_start_of_array = coords[:bl_i + 1]
bottom_edge = np.concatenate(
[coords_end_of_array, coords_start_of_array], axis=0)
yield left_edge, top_edge, right_edge, bottom_edge
except Exception:
pass
raise SudokuExtractError("No suitable blob could be found.")
def iter_blob_contours(image, n=5):
original_shape = image.shape[::-1]
if max(original_shape) < 2000:
size = (500, 500)
y_scale = original_shape[0] / 500
x_scale = original_shape[1] / 500
else:
size = (1000, 1000)
y_scale = original_shape[0] / 1000
x_scale = original_shape[1] / 1000
img = resize(image, size)
bimg = gaussian_filter(img, sigma=1.0)
bimg = threshold_adaptive(bimg, 20, offset=2/255)
bimg = (-binary_erosion(bimg))
label_image = label(bimg, background=False)
label_image += 1
regions = regionprops(label_image)
regions.sort(key=attrgetter('area'), reverse=True)
iter_n = 0
for region in regions:
iter_n += 1
if iter_n > n:
break
try:
coords = get_contours(add_border(label_image == region.label,
size=label_image.shape,
border_size=1,
background_value=False))[0]
if np.linalg.norm(coords[0, :] - coords[-1, :]) > 1e-10:
raise SudokuExtractError("Not a closed contour.")
else:
coords = np.fliplr(coords[:-1, :])
top_left = sorted(coords, key=lambda x: np.linalg.norm(np.array(x)))[0]
top_right = sorted(coords, key=lambda x: np.linalg.norm(np.array(x) - [img.shape[1], 0]))[0]
bottom_left = sorted(coords, key=lambda x: np.linalg.norm(np.array(x) - [0, img.shape[0]]))[0]
bottom_right = sorted(coords, key=lambda x: np.linalg.norm(np.array(x) - [img.shape[1], img.shape[0]]))[0]
tl_i = np.argmax((coords == top_left).sum(axis=1))
tr_i = np.argmax((coords == top_right).sum(axis=1))
bl_i = np.argmax((coords == bottom_left).sum(axis=1))
br_i = np.argmax((coords == bottom_right).sum(axis=1))
coords[:, 0] *= y_scale
coords[:, 1] *= x_scale
if tl_i > bl_i:
left_edge = coords[bl_i:tl_i + 1, :]
else:
coords_end_of_array = coords[bl_i:, :]
coords_start_of_array = coords[:tl_i + 1]
left_edge = np.concatenate([coords_end_of_array, coords_start_of_array], axis=0)
if tr_i > tl_i:
top_edge = coords[tl_i:tr_i + 1, :]
else:
coords_end_of_array = coords[tl_i:, :]
coords_start_of_array = coords[:tr_i + 1]
top_edge = np.concatenate([coords_end_of_array, coords_start_of_array], axis=0)
if br_i > tr_i:
right_edge = coords[tr_i:br_i + 1, :]
else:
coords_end_of_array = coords[tr_i:, :]
coords_start_of_array = coords[:br_i + 1]
right_edge = np.concatenate([coords_end_of_array, coords_start_of_array], axis=0)
if bl_i > br_i:
bottom_edge = coords[br_i:bl_i + 1, :]
else:
coords_end_of_array = coords[br_i:, :]
coords_start_of_array = coords[:bl_i + 1]
bottom_edge = np.concatenate([coords_end_of_array, coords_start_of_array], axis=0)
yield left_edge, top_edge, right_edge, bottom_edge
except Exception:
pass
raise SudokuExtractError("No suitable blob could be found.")