def test_ellipse_trivial():
img = np.zeros((2, 2), "uint8")
rr, cc = ellipse(0.5, 0.5, 0.5, 0.5)
img[rr, cc] = 1
img_correct = np.array([[0, 0], [0, 0]])
assert_array_equal(img, img_correct)
img = np.zeros((2, 2), "uint8")
rr, cc = ellipse(0.5, 0.5, 1.1, 1.1)
img[rr, cc] = 1
img_correct = np.array([[1, 1], [1, 1]])
assert_array_equal(img, img_correct)
img = np.zeros((3, 3), "uint8")
rr, cc = ellipse(1, 1, 0.9, 0.9)
img[rr, cc] = 1
img_correct = np.array([[0, 0, 0], [0, 1, 0], [0, 0, 0]])
assert_array_equal(img, img_correct)
img = np.zeros((3, 3), "uint8")
rr, cc = ellipse(1, 1, 1.1, 1.1)
img[rr, cc] = 1
img_correct = np.array([[0, 1, 0], [1, 1, 1], [0, 1, 0]])
assert_array_equal(img, img_correct)
img = np.zeros((3, 3), "uint8")
rr, cc = ellipse(1, 1, 1.5, 1.5)
img[rr, cc] = 1
img_correct = np.array([[1, 1, 1], [1, 1, 1], [1, 1, 1]])
assert_array_equal(img, img_correct)
def test_ellipse_rotation_symmetry():
img1 = np.zeros((150, 150), dtype=np.uint8)
img2 = np.zeros((150, 150), dtype=np.uint8)
for angle in range(0, 180, 15):
img1.fill(0)
rr, cc = ellipse(80, 70, 60, 40, rotation=np.deg2rad(angle))
img1[rr, cc] = 1
img2.fill(0)
rr, cc = ellipse(80, 70, 60, 40, rotation=np.deg2rad(angle + 180))
img2[rr, cc] = 1
assert_array_equal(img1, img2)
def add_Ellipse(self, scale = 1, thick = 5):
"""h for vertical， w for horizontal"""
h = self.e_h = self.base_e_h * scale
w = self.e_w = self.base_e_w * scale
# e and lv should not be too close
shift = self.e_w/8
center = self.lv_center + self.lv_radius_vec
center[1] += shift
self.e_center = center
self.e_thick = thick
self.e_big = d.ellipse(center[0], center[1], h, w)
self.e_small= d.ellipse(center[0], center[1], h-thick, w-thick)
def makeFakeVessels(imgsize=(512, 512), background=230):
"""
create and save a matrix with whitish background and randomly selected vessel sizes and save matrices generated as images of format png
"""
nVes = 20
mu = 20
sigma = 5
minw = 5
sx, sy = imgsize
vasc = np.ones((sx, sy), dtype=np.uint8) * background
for i in range(nVes):
cx, cy = random.uniform(0, sx), random.uniform(0, sy)
r1, r2 = 0, 0
while (r1 < minw) or (r2 < minw):
np.random.seed(20)
r1 = np.random.normal(mu, sigma)
r2 = np.random.normal(mu, sigma)
print(r1, r2)
rr, cc = ellipse(cy, cx, r1, r2)
if np.any(rr >= sy):
ix = rr < sy
rr, cc = rr[ix], cc[ix]
if np.any(cc >= sx):
ix = cc < sx
rr, cc = rr[ix], cc[ix]
vasc[rr, cc] = 1 # make circle blackish
return vasc
def test_ellipse():
img = np.zeros((15, 15), 'uint8')
rr, cc = ellipse(7, 7, 3, 7)
img[rr, cc] = 1
img_ = np.array(
[[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0],
[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0],
[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0],
[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0],
[0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]
)
assert_array_equal(img, img_)
def test_ellipse_with_shape():
img = np.zeros((15, 15), 'uint8')
rr, cc = ellipse(7, 7, 3, 10, shape=img.shape)
img[rr, cc] = 1
img_ = np.array(
[[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]
)
assert_array_equal(img, img_)
def test_ellipse_negative():
rr, cc = ellipse(-3, -3, 1.7, 1.7)
rr_, cc_ = np.nonzero(
np.array([[0, 0, 0, 0, 0], [0, 1, 1, 1, 0], [0, 1, 1, 1, 0], [0, 1, 1, 1, 0], [0, 0, 0, 0, 0]])
)
assert_array_equal(rr, rr_ - 5)
assert_array_equal(cc, cc_ - 5)
def setup():
image = np.zeros((30, 30, 3), dtype=float)
image[draw.ellipse(15, 15, 5, 8)] = 1
state = np.zeros((30, 30, 2))
state[15, 15] = (1, 1)
state[0, 0] = (0, 1)
return image, state
def get_indices(self):
"""Returns a set of points that lie inside the picked polygon."""
if not self.center:
raise ValueError("Cannot get ellipse indices before the dimensions are defined")
x, y = self.center
w = self.width
h = self.height
return ellipse(y, x, h/2., w/2., self.im)
def test_ellipse_rotated():
img = np.zeros((1000, 1200), dtype=np.uint8)
for rot in range(0, 180, 10):
img.fill(0)
angle = np.deg2rad(rot)
rr, cc = ellipse(500, 600, 200, 400, rotation=angle)
img[rr, cc] = 1
# estimate orientation of ellipse
angle_estim = np.round(regionprops(img)[0].orientation, 3) % (np.pi / 2)
assert_almost_equal(angle_estim, angle % (np.pi / 2), 2)
def test_rotate_largest_region(self):
test_img = np.zeros((1000,1000), dtype=np.uint8)
rr, cc = circle(100,100,50)
test_img[rr,cc] = 1
rr, cc = ellipse(500,500,300,100)
test_img[rr, cc] = 1
rotated = gzapi.rotate_largest_region(test_img)
largest_region = gzapi.get_largest_region(rotated)
self.assertAlmostEqual(largest_region.orientation, 0)
self.assertIsNone(gzapi.rotate_largest_region(None))
def sim_wire(angle, gaussian_sigma=1, noise_level=0, L=100):
"Draw a wire with blur and optional noise."
# Noise level should be from 0 to 1. 0.02 is reasonable.
shape = (L, L)
a = np.zeros(shape, dtype=np.uint8)
a[draw.ellipse(L//2, L//2, L//24, L//4)] = 100 # horizontal ellipse
a = filter.gaussian_filter(a, gaussian_sigma)
b = transform.rotate(a, angle)
b += noise_level*np.random.randn(*shape)
return b
def _measure_fluorescence(self, time_period, time_index, image_slice, channel_annotation):
mask = np.zeros((image_slice.height * 2, image_slice.width))
old_pole, new_pole = channel_annotation.get_cell_bounds(time_period, time_index)
ellipse_minor_radius = int(0.80 * image_slice.height)
ellipse_major_radius = int((new_pole - old_pole) / 2.0) * 0.8
centroid_y = int(image_slice.height)
centroid_x = int((new_pole + old_pole) / 2.0)
rr, cc = draw.ellipse(centroid_y, centroid_x, ellipse_minor_radius, ellipse_major_radius)
mask[rr, cc] = 1
mean, stddev, median, area, centroid = self._calculate_cell_intensity_statistics(mask.astype("int"), image_slice.image_data)
return mean, stddev, median, area, centroid
def test_growcut_basic():
image = np.zeros((30, 30, 3), dtype=float)
image[draw.ellipse(15, 15, 5, 8)] = 1
state = np.zeros((30, 30, 2))
state[15, 15] = (1, 1)
state[0, 0] = (0, 1)
npt.assert_array_equal(image[..., 0],
growcut(image, state, window_size=3, max_iter=20))
npt.assert_array_equal(image[..., 0],
growcut_fast(image, state, window_size=3, max_iter=20))
def ellipse(width, height, dtype=np.uint8):
"""Generates a flat, ellipse-shaped structuring element.
Every pixel along the perimeter of ellipse satisfies
the equation ``(x/width+1)**2 + (y/height+1)**2 = 1``.
Parameters
----------
width : int
The width of the ellipse-shaped structuring element.
height : int
The height of the ellipse-shaped structuring element.
Other Parameters
----------------
dtype : data-type
The data type of the structuring element.
Returns
-------
selem : ndarray
The structuring element where elements of the neighborhood
are 1 and 0 otherwise.
Examples
--------
>>> from skimage.morphology import selem
>>> selem.ellipse(5, 3)
array([[0, 0, 1, 1, 1, 1, 1, 1, 1, 0, 0],
[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
[0, 0, 1, 1, 1, 1, 1, 1, 1, 0, 0]], dtype=uint8)
"""
selem = np.zeros((2 * height + 1, 2 * width + 1), dtype=dtype)
rows, cols = draw.ellipse(height, width, height + 1, width + 1)
selem[rows, cols] = 1
return selem
def ellipseMask(maskImg, ellipse_yradius, ellipse_xradius):
boxsize = maskImg.get_xsize()
maskArray = EMNumPy.em2numpy(maskImg)
if (boxsize <= ellipse_yradius * 2 or boxsize <= ellipse_xradius * 2):
print "ERROR: ellipse_xradius or ellipse_yradius is larger than the boxsize of particles."
sys.exit()
#from skimage.draw import ellipse
#Generate coordinates of pixels within ellipse.
####################################
'''
ellipse_xradius=random.randint(ellipse_xradius-10,ellipse_xradius+10)
print ellipse_xradius
ellipse_yradius=random.randint(ellipse_yradius-10,ellipse_yradius+10)
print ellipse_yradius
'''
size = max([ellipse_xradius, ellipse_yradius]) * 2 + 4
cx = size/2
cy = size/2
ellipseArray = np.zeros((size, size), dtype=np.uint8)
rr, cc = ellipse(cy, cx, ellipse_yradius, ellipse_xradius)
ellipseArray[rr, cc] = 1
m, n = ellipseArray.shape
assert m==n
if (m%2 == 0):
pad_before = (boxsize - m)/2
pad_after = (boxsize - m)/2
else:
pad_before = (boxsize - m)/2
pad_after = (boxsize - m)/2+1
ellipseArrayPad = np.pad(ellipseArray, (pad_before, pad_after), mode='constant')
ellipseImg = EMNumPy.numpy2em(ellipseArrayPad)
return ellipseImg
import matplotlib.pyplot as plt import numpy as np from skimage import measure from skimage.draw import ellipse from skimage.transform import rotate image = np.zeros((600, 600)) rr, cc = ellipse(300, 350, 100, 220) image[rr, cc] = 1 image = rotate(image, angle=15, order=0) label_img = measure.label(input=image) print(label_img.shape, label_img.max(), label_img.min(), np.unique(label_img)) print(np.sum(image - label_img)) # regions = measure.regionprops(label_image=label_img) print(regions) fig, ax = plt.subplots() ax.imshow(image, cmap=plt.cm.gray) # pylint: disable=E1101
def read_roi(roi_obj):
"""Parses an individual ImageJ ROI
_getX lines with no assignment are bytes within the imageJ roi file
format that contain additional information that can be extracted if
needed. In line comments label what they are.
This is based on:
http://rsbweb.nih.gov/ij/developer/source/ij/io/RoiDecoder.java.html
http://rsbweb.nih.gov/ij/developer/source/ij/io/RoiEncoder.java.html
Parameters
----------
roi_obj : file object
File object containing a single ImageJ ROI
Returns
-------
ROI
Returns a parsed ROI object (dictionary)
Raises
------
IOError
If there is an error reading the roi file object
ValueError
If unable to parse ROI
"""
sub_pixel_resolution = 128
# Other options that are not currently used
# SPLINE_FIT = 1
# DOUBLE_HEADED = 2
# OUTLINE = 4
# OVERLAY_LABELS = 8
# OVERLAY_NAMES = 16
# OVERLAY_BACKGROUNDS = 32
# OVERLAY_BOLD = 64
# DRAW_OFFSET = 256
pos = [4]
def _get8():
"""Read 1 byte from the roi file object"""
pos[0] += 1
s = roi_obj.read(1)
if not s:
raise IOError('read_imagej_roi: Unexpected EOF')
return ord(s)
def _get16():
"""Read 2 bytes from the roi file object"""
b0 = _get8()
b1 = _get8()
return (b0 << 8) | b1
def _get16signed():
"""Read a signed 16 bit integer from 2 bytes from roi file object"""
b0 = _get8()
b1 = _get8()
out = (b0 << 8) | b1
# This is a signed integer, so need to check if the value is
# positive or negative.
if b0 > 127:
out = out - 65536
return out
def _get32():
"""Read 4 bytes from the roi file object"""
s0 = _get16()
s1 = _get16()
return (s0 << 16) | s1
def _getfloat():
"""Read a float from the roi file object"""
v = np.int32(_get32())
return v.view(np.float32)
def _getcoords(z=0):
"""Get the next coordinate of an roi polygon"""
if options & sub_pixel_resolution:
getc = _getfloat
points = np.empty((n_coordinates, 3), dtype=np.float32)
else:
getc = _get16
points = np.empty((n_coordinates, 3), dtype=np.int16)
points[:, 0] = [getc() for _ in range(n_coordinates)]
points[:, 1] = [getc() for _ in range(n_coordinates)]
points[:, 0] += left
points[:, 1] += top
points[:, 2] = z
return points
magic = roi_obj.read(4)
if magic != b'Iout':
raise IOError('read_imagej_roi: Magic number not found')
_get16() # version
roi_type = _get8()
# Discard extra second Byte:
_get8()
if not (0 <= roi_type < 11):
raise ValueError('read_imagej_roi: \
ROI type {} not supported'.format(roi_type))
top = _get16signed()
left = _get16signed()
bottom = _get16signed()
right = _get16signed()
n_coordinates = _get16()
x1 = _getfloat() # x1
y1 = _getfloat() # y1
x2 = _getfloat() # x2
y2 = _getfloat() # y2
_get16() # stroke width
_get32() # shape roi size
_get32() # stroke color
_get32() # fill color
subtype = _get16()
if subtype != 0 and subtype != 3:
raise ValueError('read_imagej_roi: \
ROI subtype {} not supported (!= 0)'.format(subtype))
options = _get16()
if subtype == 3 and roi_type == 7:
# ellipse aspect ratio
aspect_ratio = _getfloat()
else:
_get8() # arrow style
_get8() # arrow head size
_get16() # rectangle arc size
z = _get32() # position
if z > 0:
z -= 1 # Multi-plane images start indexing at 1 instead of 0
_get32() # header 2 offset
if roi_type == 0:
# Polygon
coords = _getcoords(z)
coords = coords.astype('float')
return {'polygons': coords}
elif roi_type == 1:
# Rectangle
coords = [[left, top, z], [right, top, z], [right, bottom, z],
[left, bottom, z]]
coords = np.array(coords).astype('float')
return {'polygons': coords}
elif roi_type == 2:
# Oval
width = right - left
height = bottom - top
# 0.5 moves the mid point to the center of the pixel
x_mid = (right + left) / 2.0 - 0.5
y_mid = (top + bottom) / 2.0 - 0.5
mask = np.zeros((z + 1, right, bottom), dtype=bool)
for y, x in product(np.arange(top, bottom), np.arange(left, right)):
mask[z, x, y] = ((x - x_mid) ** 2 / (width / 2.0) ** 2 +
(y - y_mid) ** 2 / (height / 2.0) ** 2 <= 1)
return {'mask': mask}
elif roi_type == 7:
if subtype == 3:
# ellipse
mask = np.zeros((1, right+10, bottom+10), dtype=bool)
r_radius = np.sqrt((x2-x1)**2+(y2-y1)**2)/2.0
c_radius = r_radius*aspect_ratio
r = (x1+x2)/2-0.5
c = (y1+y2)/2-0.5
shpe = mask.shape
orientation = np.arctan2(y2-y1, x2-x1)
X, Y = ellipse(r, c, r_radius, c_radius, shpe[1:], orientation)
mask[0, X, Y] = True
return {'mask': mask}
else:
# Freehand
coords = _getcoords(z)
coords = coords.astype('float')
return {'polygons': coords}
else:
try:
coords = _getcoords(z)
coords = coords.astype('float')
return {'polygons': coords}
except BaseException:
raise ValueError(
'read_imagej_roi: ROI type {} not supported'.format(roi_type))
def bp_removal_2d(obj_tmp, cy, cx, fwhm, sig, protect_psf, verbose):
n_x = obj_tmp.shape[1]
n_y = obj_tmp.shape[0]
# Squash the frame if twice less resolved vertically than horizontally
if half_res_y:
if n_y % 2 != 0:
msg = 'The input frames do not have of an even number of rows. '
msg2 = 'Hence, you should not use option half_res_y = True'
raise ValueError(msg + msg2)
n_y = int(n_y / 2)
cy = int(cy / 2)
frame = obj_tmp.copy()
obj_tmp = np.zeros([n_y, n_x])
for yy in range(n_y):
obj_tmp[yy] = frame[2 * yy]
#1/ Stddev of background
_, _, stddev = sigma_clipped_stats(obj_tmp, sigma=2.5)
#2/ Define each annulus, its median and stddev
ymax = max(cy, n_y - cy)
xmax = max(cx, n_x - cx)
if half_res_y:
ymax *= 2
rmax = np.sqrt(ymax**2 + xmax**2)
# the annuli definition is optimized for Airy rings
ann_width = max(1.5, 0.61 * fwhm)
nrad = int(rmax / ann_width) + 1
d_bord_max = max(n_y - cy, cy, n_x - cx, cx)
if half_res_y:
d_bord_max = max(2 * (n_y - cy), 2 * cy, n_x - cx, cx)
big_ell_frame = np.zeros_like(obj_tmp)
sma_ell_frame = np.zeros_like(obj_tmp)
ann_frame_cumul = np.zeros_like(obj_tmp)
n_neig = np.zeros(nrad)
med_neig = np.zeros(nrad)
std_neig = np.zeros(nrad)
neighbours = np.zeros([nrad, n_y * n_x])
for rr in range(nrad):
if rr > int(d_bord_max / ann_width):
# just to merge farthest annuli with very few elements
rr_big = nrad
rr_sma = int(d_bord_max / ann_width)
else:
rr_big = rr
rr_sma = rr
if half_res_y:
big_ell_idx = ellipse(cy=cy,
cx=cx,
yradius=((rr_big + 1) * ann_width) / 2,
xradius=(rr_big + 1) * ann_width,
shape=(n_y, n_x))
if rr != 0:
small_ell_idx = ellipse(cy=cy,
cx=cx,
yradius=(rr_sma * ann_width) / 2,
xradius=rr_sma * ann_width,
shape=(n_y, n_x))
else:
big_ell_idx = circle(cy,
cx,
radius=(rr_big + 1) * ann_width,
shape=(n_y, n_x))
if rr != 0:
small_ell_idx = circle(cy,
cx,
radius=rr_sma * ann_width,
shape=(n_y, n_x))
big_ell_frame[big_ell_idx] = 1
if rr != 0: sma_ell_frame[small_ell_idx] = 1
ann_frame = big_ell_frame - sma_ell_frame
n_neig[rr] = ann_frame[np.where(ann_frame)].shape[0]
neighbours[rr, :n_neig[rr]] = obj_tmp[np.where(ann_frame)]
ann_frame_cumul[np.where(ann_frame)] = rr
# We delete iteratively max and min outliers in each annulus,
# so that the annuli median and stddev are not corrupted by bpixs
neigh = neighbours[rr, :n_neig[rr]]
n_removal = 0
n_pix_init = neigh.shape[0]
while neigh.shape[0] > 10 and n_removal < n_pix_init / 10:
min_neigh = np.amin(neigh)
if reject_outliers(neigh,
min_neigh,
m=5,
stddev=stddev,
min_thr=min_thr * stddev,
mid_thr=mid_thr * stddev):
min_idx = np.argmin(neigh)
neigh = np.delete(neigh, min_idx)
n_removal += 1
else:
max_neigh = np.amax(neigh)
if reject_outliers(neigh,
max_neigh,
m=5,
stddev=stddev,
min_thr=min_thr * stddev,
mid_thr=mid_thr * stddev):
max_idx = np.argmax(neigh)
neigh = np.delete(neigh, max_idx)
n_removal += 1
else:
break
n_neig[rr] = neigh.shape[0]
neighbours[rr, :n_neig[rr]] = neigh
neighbours[rr, n_neig[rr]:] = 0
med_neig[rr] = np.median(neigh)
std_neig[rr] = np.std(neigh)
#3/ Create a tuple-array with coordinates of a circle of radius 1.8*fwhm
# centered on the provided coordinates of the star
if protect_psf:
if half_res_y:
circl_new = ellipse(cy,
cx,
yradius=0.9 * fwhm,
xradius=1.8 * fwhm,
shape=(n_y, n_x))
else:
circl_new = circle(cy, cx, radius=1.8 * fwhm, shape=(n_y, n_x))
else:
circl_new = []
#4/ Loop on all pixels to check bpix
bpix_map = np.zeros_like(obj_tmp)
obj_tmp_corr = obj_tmp.copy()
for yy in range(n_y):
for xx in range(n_x):
if half_res_y:
rad = np.sqrt((2 * (cy - yy))**2 + (cx - xx)**2)
else:
rad = dist(cy, cx, yy, xx)
rr = int(rad / ann_width)
neigh = neighbours[rr, :n_neig[rr]]
dev = max(stddev, min(std_neig[rr], med_neig[rr]))
# check min_thr
if obj_tmp[yy, xx] < min_thr * stddev:
bpix_map[yy, xx] = 1
# Gaussian noise
#obj_tmp_corr[yy,xx] = med_neig[rr] +dev*np.random.randn()
# Poisson noise
rand_fac = 2 * (np.random.rand() - 0.5)
obj_tmp_corr[yy,xx] = med_neig[rr] + \
np.sqrt(np.abs(med_neig[rr]))*rand_fac
# check median +- sig*stddev
elif (obj_tmp[yy, xx] < med_neig[rr] - sig * dev
or obj_tmp[yy, xx] > med_neig[rr] + sig * dev):
bpix_map[yy, xx] = 1
# Gaussian noise
#obj_tmp_corr[yy,xx] = med_neig[rr] +dev*np.random.randn()
# Poisson noise
rand_fac = 2 * (np.random.rand() - 0.5)
obj_tmp_corr[yy,xx] = med_neig[rr] + \
np.sqrt(np.abs(med_neig[rr]))*rand_fac
# check mid_thr and neighbours
else:
min_el = max(2, 0.05 * n_neig[rr])
if (obj_tmp[yy, xx] < mid_thr * stddev
and neigh[neigh < (mid_thr + 5) * stddev].shape[0]
< min_el):
bpix_map[yy, xx] = 1
# Gaussian noise
#obj_tmp_corr[yy,xx] = med_neig[rr] + \
# dev*np.random.randn()
# Poisson noise
rand_fac = 2 * (np.random.rand() - 0.5)
obj_tmp_corr[yy,xx] = med_neig[rr] + \
np.sqrt(np.abs(med_neig[rr]))*rand_fac
#5/ Count bpix and uncorrect if within the circle
nbpix_tot = np.sum(bpix_map)
nbpix_tbc = nbpix_tot - np.sum(bpix_map[circl_new])
bpix_map[circl_new] = 0
obj_tmp_corr[circl_new] = obj_tmp[circl_new]
if verbose:
print(nbpix_tot, ' bpix in total, and ', nbpix_tbc, ' corrected.')
# Unsquash all the frames
if half_res_y:
frame = obj_tmp_corr.copy()
frame_bpix = bpix_map.copy()
n_y = 2 * n_y
obj_tmp_corr = np.zeros([n_y, n_x])
bpix_map = np.zeros([n_y, n_x])
ann_frame = ann_frame_cumul.copy()
ann_frame_cumul = np.zeros([n_y, n_x])
for yy in range(n_y):
obj_tmp_corr[yy] = frame[int(yy / 2)]
bpix_map[yy] = frame_bpix[int(yy / 2)]
ann_frame_cumul[yy] = ann_frame[int(yy / 2)]
return obj_tmp_corr, bpix_map, ann_frame_cumul
def add_ellipse(self, xy_c, wx, wy, idx=1):
rr, cc = ellipse(*xy_c, wx, wy, shape=self.mask.shape)
self.mask[rr, cc] = idx
def bp_removal_2d(obj_tmp, cy, cx, fwhm, sig, protect_psf, verbose):
n_x = obj_tmp.shape[1]
n_y = obj_tmp.shape[0]
# Squash the frame if twice less resolved vertically than horizontally
if half_res_y:
if n_y % 2 != 0:
msg = 'The input frames do not have of an even number of rows. '
msg2 = 'Hence, you should not use option half_res_y = True'
raise ValueError(msg+msg2)
n_y = int(n_y/2)
cy = int(cy/2)
frame = obj_tmp.copy()
obj_tmp = np.zeros([n_y,n_x])
for yy in range(n_y):
obj_tmp[yy] = frame[2*yy]
#1/ Stddev of background
_, _, stddev = sigma_clipped_stats(obj_tmp, sigma=2.5)
#2/ Define each annulus, its median and stddev
ymax = max(cy, n_y-cy)
xmax = max(cx, n_x-cx)
if half_res_y:
ymax *= 2
rmax = np.sqrt(ymax**2+xmax**2)
# the annuli definition is optimized for Airy rings
ann_width = max(1.5, 0.61*fwhm)
nrad = int(rmax/ann_width)+1
d_bord_max = max(n_y-cy, cy, n_x-cx, cx)
if half_res_y:
d_bord_max = max(2*(n_y-cy), 2*cy, n_x-cx, cx)
big_ell_frame = np.zeros_like(obj_tmp)
sma_ell_frame = np.zeros_like(obj_tmp)
ann_frame_cumul = np.zeros_like(obj_tmp)
n_neig = np.zeros(nrad)
med_neig = np.zeros(nrad)
std_neig = np.zeros(nrad)
neighbours = np.zeros([nrad,n_y*n_x])
for rr in range(nrad):
if rr > int(d_bord_max/ann_width):
# just to merge farthest annuli with very few elements
rr_big = nrad
rr_sma = int(d_bord_max/ann_width)
else:
rr_big = rr
rr_sma= rr
if half_res_y:
big_ell_idx = ellipse(cy=cy, cx=cx,
yradius=((rr_big+1)*ann_width)/2,
xradius=(rr_big+1)*ann_width,
shape=(n_y,n_x))
if rr != 0:
small_ell_idx = ellipse(cy=cy, cx=cx,
yradius=(rr_sma*ann_width)/2,
xradius=rr_sma*ann_width,
shape=(n_y,n_x))
else:
big_ell_idx = circle(cy, cx, radius=(rr_big+1)*ann_width,
shape=(n_y,n_x))
if rr != 0:
small_ell_idx = circle(cy, cx, radius=rr_sma*ann_width,
shape=(n_y,n_x))
big_ell_frame[big_ell_idx] = 1
if rr!=0: sma_ell_frame[small_ell_idx] = 1
ann_frame = big_ell_frame - sma_ell_frame
n_neig[rr] = ann_frame[np.where(ann_frame)].shape[0]
neighbours[rr,:n_neig[rr]] = obj_tmp[np.where(ann_frame)]
ann_frame_cumul[np.where(ann_frame)] = rr
# We delete iteratively max and min outliers in each annulus,
# so that the annuli median and stddev are not corrupted by bpixs
neigh = neighbours[rr,:n_neig[rr]]
n_removal = 0
n_pix_init = neigh.shape[0]
while neigh.shape[0] > 10 and n_removal < n_pix_init/10:
min_neigh = np.amin(neigh)
if reject_outliers(neigh, min_neigh, m=5, stddev=stddev,
min_thr=min_thr*stddev,
mid_thr=mid_thr*stddev):
min_idx = np.argmin(neigh)
neigh = np.delete(neigh,min_idx)
n_removal += 1
else:
max_neigh = np.amax(neigh)
if reject_outliers(neigh, max_neigh, m=5, stddev=stddev,
min_thr=min_thr*stddev,
mid_thr=mid_thr*stddev):
max_idx = np.argmax(neigh)
neigh = np.delete(neigh,max_idx)
n_removal += 1
else: break
n_neig[rr] = neigh.shape[0]
neighbours[rr,:n_neig[rr]] = neigh
neighbours[rr,n_neig[rr]:] = 0
med_neig[rr] = np.median(neigh)
std_neig[rr] = np.std(neigh)
#3/ Create a tuple-array with coordinates of a circle of radius 1.8*fwhm
# centered on the provided coordinates of the star
if protect_psf:
if half_res_y:
circl_new = ellipse(cy, cx, yradius=0.9*fwhm,
xradius=1.8*fwhm, shape=(n_y,n_x))
else:
circl_new = circle(cy, cx, radius=1.8*fwhm,
shape=(n_y, n_x))
else: circl_new = []
#4/ Loop on all pixels to check bpix
bpix_map = np.zeros_like(obj_tmp)
obj_tmp_corr = obj_tmp.copy()
for yy in range(n_y):
for xx in range(n_x):
if half_res_y:
rad = np.sqrt((2*(cy-yy))**2+(cx-xx)**2)
else:
rad = dist(cy, cx, yy, xx)
rr = int(rad/ann_width)
neigh = neighbours[rr,:n_neig[rr]]
dev = max(stddev,min(std_neig[rr],med_neig[rr]))
# check min_thr
if obj_tmp[yy,xx] < min_thr*stddev:
bpix_map[yy,xx] = 1
# Gaussian noise
#obj_tmp_corr[yy,xx] = med_neig[rr] +dev*np.random.randn()
# Poisson noise
rand_fac= 2*(np.random.rand()-0.5)
obj_tmp_corr[yy,xx] = med_neig[rr] + \
np.sqrt(np.abs(med_neig[rr]))*rand_fac
# check median +- sig*stddev
elif (obj_tmp[yy,xx] < med_neig[rr]-sig*dev or
obj_tmp[yy,xx] > med_neig[rr]+sig*dev):
bpix_map[yy,xx] = 1
# Gaussian noise
#obj_tmp_corr[yy,xx] = med_neig[rr] +dev*np.random.randn()
# Poisson noise
rand_fac= 2*(np.random.rand()-0.5)
obj_tmp_corr[yy,xx] = med_neig[rr] + \
np.sqrt(np.abs(med_neig[rr]))*rand_fac
# check mid_thr and neighbours
else:
min_el = max(2, 0.05*n_neig[rr])
if (obj_tmp[yy,xx] < mid_thr*stddev and
neigh[neigh<(mid_thr+5)*stddev].shape[0] < min_el):
bpix_map[yy,xx] = 1
# Gaussian noise
#obj_tmp_corr[yy,xx] = med_neig[rr] + \
# dev*np.random.randn()
# Poisson noise
rand_fac = 2*(np.random.rand()-0.5)
obj_tmp_corr[yy,xx] = med_neig[rr] + \
np.sqrt(np.abs(med_neig[rr]))*rand_fac
#5/ Count bpix and uncorrect if within the circle
nbpix_tot = np.sum(bpix_map)
nbpix_tbc = nbpix_tot - np.sum(bpix_map[circl_new])
bpix_map[circl_new] = 0
obj_tmp_corr[circl_new] = obj_tmp[circl_new]
if verbose:
print(nbpix_tot, ' bpix in total, and ', nbpix_tbc, ' corrected.')
# Unsquash all the frames
if half_res_y:
frame = obj_tmp_corr.copy()
frame_bpix = bpix_map.copy()
n_y = 2*n_y
obj_tmp_corr = np.zeros([n_y,n_x])
bpix_map = np.zeros([n_y,n_x])
ann_frame = ann_frame_cumul.copy()
ann_frame_cumul = np.zeros([n_y,n_x])
for yy in range(n_y):
obj_tmp_corr[yy] = frame[int(yy/2)]
bpix_map[yy] = frame_bpix[int(yy/2)]
ann_frame_cumul[yy] = ann_frame[int(yy/2)]
return obj_tmp_corr, bpix_map, ann_frame_cumul
def bp_removal_2d(obj_tmp, cy, cx, fwhm, sig, protect_psf, verbose):
n_x = obj_tmp.shape[1]
n_y = obj_tmp.shape[0]
if half_res_y:
if n_y%2 != 0:
msg = 'The input frames do not have of an even number of rows. '
msg2 = 'Hence, you should not use option half_res_y = True'
raise ValueError(msg+msg2)
n_y = int(n_y/2)
frame = obj_tmp.copy()
obj_tmp = np.zeros([n_y,n_x])
for yy in range(n_y):
obj_tmp[yy] = frame[2*yy]
fwhm_round = int(round(fwhm))
# This should reduce the chance to accidently correct a bright planet:
if fwhm_round % 2 == 0:
neighbor_box = max(3, fwhm_round+1)
else:
neighbor_box = max(3, fwhm_round)
nneig = sum(np.arange(3, neighbor_box+2, 2))
#1/ Create a tuple-array with coordinates of a circle of radius 1.8*fwhm
# centered on the approximate coordinates of the star
if protect_psf:
if half_res_y:
circl_new = ellipse(int(cy/2), cx, yradius=0.9*fwhm,
xradius=1.8*fwhm, shape=(n_y, n_x))
else: circl_new = circle(cy, cx, radius=1.8*fwhm,
shape=(n_y, n_x))
else: circl_new = []
#2/ Compute stddev of background
# empirically 2.5 is best to find stddev of background noise
_, _, stddev = sigma_clipped_stats(obj_tmp, sigma=2.5)
#3/ Create a bad pixel map, by detecting them with find_outliers
bpix_map = find_outliers(obj_tmp, sig_dist=sig, stddev=stddev,
neighbor_box=neighbor_box,min_thr=min_thr,
mid_thr=mid_thr)
nbpix_tot = np.sum(bpix_map)
nbpix_tbc = nbpix_tot - np.sum(bpix_map[circl_new])
bpix_map[circl_new] = 0
bpix_map_cumul = np.zeros_like(bpix_map)
bpix_map_cumul[:] = bpix_map[:]
nit = 0
#4/ Loop over the bpix correction with sigma_filter, until 0 bpix left
while nbpix_tbc > 0 and nit < max_nit:
nit = nit+1
if verbose:
print("Iteration {}: {} bpix in total, {} to be "
"corrected".format(nit, nbpix_tot, nbpix_tbc))
obj_tmp = sigma_filter(obj_tmp, bpix_map, neighbor_box=neighbor_box,
min_neighbors=nneig, verbose=verbose)
bpix_map = find_outliers(obj_tmp, sig_dist=sig, in_bpix=bpix_map,
stddev=stddev, neighbor_box=neighbor_box,
min_thr=min_thr,mid_thr=mid_thr)
nbpix_tot = np.sum(bpix_map)
nbpix_tbc = nbpix_tot - np.sum(bpix_map[circl_new])
bpix_map[circl_new] = 0
bpix_map_cumul = bpix_map_cumul+bpix_map
if verbose:
print('All bad pixels are corrected.')
if half_res_y:
frame = obj_tmp.copy()
frame_bpix = bpix_map_cumul.copy()
n_y = 2*n_y
obj_tmp = np.zeros([n_y,n_x])
bpix_map_cumul = np.zeros([n_y,n_x])
for yy in range(n_y):
obj_tmp[yy] = frame[int(yy/2)]
bpix_map_cumul[yy] = frame_bpix[int(yy/2)]
return obj_tmp, bpix_map_cumul
This example shows how to measure properties of labelled image regions.
"""
import math
import matplotlib.pyplot as plt
import numpy as np
from skimage.draw import ellipse
from skimage.measure import label, regionprops
from skimage.transform import rotate
image = np.zeros((600, 600))
rr, cc = ellipse(300, 350, 100, 220)
image[rr, cc] = 1
image = rotate(image, angle=15, order=0)
label_img = label(image)
regions = regionprops(label_img)
fig, ax = plt.subplots()
ax.imshow(image, cmap=plt.cm.gray)
for props in regions:
y0, x0 = props.centroid
orientation = props.orientation
x1 = x0 + math.cos(orientation) * 0.5 * props.major_axis_length
y1 = y0 - math.sin(orientation) * 0.5 * props.major_axis_length
def draw_mask(img, percentage_x=100, percentage_y=100, offset_x=0, offset_y=0, rotation=0, rectangle=True):
"""
Convenience wrapper for `skimage.draw.ellipse` and `skimage.draw.polygon`.
Draws an ellipse at (center + ``offset``) of ``img`` with major and
minor axes as a percentage of ``img.shape``. Alternatively, draws a rectangle
centered at the offset that covers the given percentage of the height and width
of the image.
Parameters
----------
img : array
Image array that you want to draw a mask on
percentage_x : float
What percentage of the x-dimension of ``img`` do you want the x-axis to cover? Can
be greater than zero
percentage_y : float
offset_x : float
From the middle, what percentage of the width of the picture do you want the center of the
ellipse to be? Positive is left.
offset_y : float
Positive is up.
rotation : float
In [-pi, pi]. See ``skimage.draw.ellipse''
rectangel : bool
Make a rectangular mask instead of an elliptical mask?
Returns
-------
A binary array with pixels in an ellipse.
References
----------
See source and examples for `skimage.draw.ellipse` and `skimage.draw.polygon`.
"""
ydim, xdim = img.shape
mask = np.zeros((ydim, xdim))
# Convert percentages to fractions
offset_x = (xdim * offset_x/100)
offset_y = (ydim * offset_y/100)
percentage_x = percentage_x/100
percentage_y = percentage_y/100
if rectangle is False:
x_rad = np.floor((img.shape[1]/2) * percentage_x)
y_rad = np.floor((img.shape[0]/2) * percentage_y)
x_center = img.shape[1]//2 + offset_x
y_center = img.shape[0]//2 - offset_y
[x, y] = draw.ellipse(y_center, x_center, y_rad, x_rad, shape = img.shape, rotation=rotation)
else:
ysub = ydim * (1 - percentage_y)
y1 = max(ysub/2 - offset_y, 0)
y2 = min(ydim - ysub/2 - offset_y, ydim)
r_coords = np.array([y1, y1, y2, y2, y1])
xsub = xdim * (1 - percentage_x)
x1 = max(xsub/2 + offset_x,0)
x2 = min(xdim - xsub/2 + offset_x, xdim)
c_coords = np.array([x1, x2, x2, x1, x1])
x, y = draw.polygon(r_coords, c_coords)
mask[x, y] = 1
return(mask)
.. [1] http://en.wikipedia.org/wiki/Corner_detection
.. [2] http://en.wikipedia.org/wiki/Interest_point_detection
"""
from matplotlib import pyplot as plt
from skimage import data
from skimage.feature import corner_harris, corner_subpix, corner_peaks
from skimage.transform import warp, AffineTransform
from skimage.draw import ellipse
tform = AffineTransform(scale=(1.3, 1.1), rotation=1, shear=0.7,
translation=(210, 50))
image = warp(data.checkerboard(), tform.inverse, output_shape=(350, 350))
rr, cc = ellipse(310, 175, 10, 100)
image[rr, cc] = 1
image[180:230, 10:60] = 1
image[230:280, 60:110] = 1
coords = corner_peaks(corner_harris(image), min_distance=5)
coords_subpix = corner_subpix(image, coords, window_size=13)
fig, ax = plt.subplots()
ax.imshow(image, interpolation='nearest', cmap=plt.cm.gray)
ax.plot(coords[:, 1], coords[:, 0], '.b', markersize=3)
ax.plot(coords_subpix[:, 1], coords_subpix[:, 0], '+r', markersize=15)
ax.axis((0, 350, 350, 0))
plt.show()
ax3.set_axis_off()
ax3.set_title('Masked pixels')
plt.show()
##################################################
# Solid masks are another illustrating example. In this case, we have a
# limited view of an image and an offset image. The masks for these images
# need not be the same. The `masked_register_translation` function will correctly identify
# which part of the images should be compared.
image = data.camera()
shift = (-22, 13)
rr1, cc1 = draw.ellipse(259, 248,
r_radius = 125, c_radius = 100,
shape = image.shape)
rr2, cc2 = draw.ellipse(300, 200,
r_radius = 110, c_radius = 180,
shape = image.shape)
mask1 = np.zeros_like(image, dtype = np.bool)
mask2 = np.zeros_like(image, dtype = np.bool)
mask1[rr1, cc1] = True
mask2[rr2, cc2] = True
offset_image = ndi.shift(image, shift)
image *= mask1
offset_image *= mask2
def draw_ellipse(img, center, axis):
rr, cc = draw.ellipse(center[0], center[1], axis[0], axis[1],
shape=img.shape)
img[rr, cc] = 1
return img
def axis_likelyhood_surface(data, kernel_width = 5, center_margin = 8, num_peaks = 10, num_circles = 20, upscale = 1.5, minradius = 15, maxradius = 65, radstep = 2):
ximagesize=data[0]['data'].shape[1]
yimagesize=data[0]['data'].shape[2]
maxradius = max(int(maxradius / np.max(data[0]['metadata']['PixelSpacing'])),1)
minradius = max(int(minradius / np.max(data[0]['metadata']['PixelSpacing'])),1)
radstep = max(int(radstep / np.max(data[0]['metadata']['PixelSpacing'])),1)
xsurface=np.tile(range(ximagesize),(yimagesize,1)).T
ysurface=np.tile(range(yimagesize),(ximagesize,1))
lsurface=np.zeros((ximagesize,yimagesize))
allcenters = []
allaccums = []
allradii = []
for ddi in data:
outdata=ddi['data'].copy()
ff1 = fftn(outdata)
fh = np.absolute(ifftn(ff1[1, :, :]))
fh[fh < 0.1* np.max(fh)] =0.0
image=img_as_ubyte(fh/np.max(fh))
#image = fh/np.max(fh)
# find hough circles
edges = canny(image, sigma=3, low_threshold=10, high_threshold=50)
#edges = canny(image)
# Detect two radii
hough_radii = np.arange(minradius, maxradius, radstep)
hough_res = hough_circle(edges, hough_radii)
if hough_res.any():
centers = []
accums = []
radii = []
for radius, h in zip(hough_radii, hough_res):
# For each radius, extract num_peaks circles
peaks = peak_local_max(h, num_peaks=num_peaks)
centers.extend(peaks)
accums.extend(h[peaks[:, 0], peaks[:, 1]])
radii.extend([radius] * num_peaks)
# Keep the most prominent num_circles circles
sorted=np.argsort(accums)[::-1][:num_circles]
for idx in sorted:
center_x, center_y = centers[idx]
allcenters.append(centers[idx])
allradii.append(radii[idx])
allaccums.append(accums[idx])
brightness=accums[idx]
lsurface = lsurface + brightness* np.exp(-((xsurface-center_x)**2 + (ysurface-center_y)**2)/kernel_width**2)
lsurface=lsurface/lsurface.max()
# select most likely ROI center
x_axis, y_axis = np.unravel_index(lsurface.argmax(), lsurface.shape)
# determine ROI radius
x_radius = 0
y_radius = 0
for idx in range(len(allcenters)):
xshift=np.abs(allcenters[idx][0]-x_axis)
yshift=np.abs(allcenters[idx][1]-y_axis)
if (xshift <= center_margin) & (yshift <= center_margin):
x_radius = np.max((x_radius,allradii[idx]+xshift))
y_radius = np.max((y_radius,allradii[idx]+yshift))
x_radius = upscale * x_radius
y_radius = upscale * y_radius
#x_radius = 1.5*maxradius
#y_radius = 1.5*maxradius
ROImask = np.zeros_like(lsurface)
[rr,cc] = ellipse(x_axis, y_axis, x_radius, y_radius)
dum=(rr < ximagesize) & (rr>=0) & (cc < yimagesize) & (cc>=0)
rr=rr[dum]
cc=cc[dum]
ROImask[rr,cc]=1.
return lsurface, ROImask, (x_axis, y_axis), (x_radius, y_radius)
def bp_removal_2d(obj_tmp, cy, cx, fwhm, sig, protect_psf, verbose):
n_x = obj_tmp.shape[1]
n_y = obj_tmp.shape[0]
if half_res_y:
if n_y % 2 != 0:
msg = 'The input frames do not have of an even number of rows. '
msg2 = 'Hence, you should not use option half_res_y = True'
raise ValueError(msg + msg2)
n_y = int(n_y / 2)
frame = obj_tmp.copy()
obj_tmp = np.zeros([n_y, n_x])
for yy in range(n_y):
obj_tmp[yy] = frame[2 * yy]
fwhm_round = int(round(fwhm))
# This should reduce the chance to accidently correct a bright planet:
if fwhm_round % 2 == 0:
neighbor_box = max(3, fwhm_round + 1)
else:
neighbor_box = max(3, fwhm_round)
nneig = sum(np.arange(3, neighbor_box + 2, 2))
#1/ Create a tuple-array with coordinates of a circle of radius 1.8*fwhm
# centered on the approximate coordinates of the star
if protect_psf:
if half_res_y:
circl_new = ellipse(int(cy / 2),
cx,
yradius=0.9 * fwhm,
xradius=1.8 * fwhm,
shape=(n_y, n_x))
else:
circl_new = circle(cy, cx, radius=1.8 * fwhm, shape=(n_y, n_x))
else:
circl_new = []
#2/ Compute stddev of background
# empirically 2.5 is best to find stddev of background noise
_, _, stddev = sigma_clipped_stats(obj_tmp, sigma=2.5)
#3/ Create a bad pixel map, by detecting them with find_outliers
bpix_map = find_outliers(obj_tmp,
sig_dist=sig,
stddev=stddev,
neighbor_box=neighbor_box,
min_thr=min_thr,
mid_thr=mid_thr)
nbpix_tot = np.sum(bpix_map)
nbpix_tbc = nbpix_tot - np.sum(bpix_map[circl_new])
bpix_map[circl_new] = 0
bpix_map_cumul = np.zeros_like(bpix_map)
bpix_map_cumul[:] = bpix_map[:]
nit = 0
#4/ Loop over the bpix correction with sigma_filter, until 0 bpix left
while nbpix_tbc > 0 and nit < max_nit:
nit = nit + 1
if verbose:
print("Iteration {}: {} bpix in total, {} to be "
"corrected".format(nit, nbpix_tot, nbpix_tbc))
obj_tmp = sigma_filter(obj_tmp,
bpix_map,
neighbor_box=neighbor_box,
min_neighbors=nneig,
verbose=verbose)
bpix_map = find_outliers(obj_tmp,
sig_dist=sig,
in_bpix=bpix_map,
stddev=stddev,
neighbor_box=neighbor_box,
min_thr=min_thr,
mid_thr=mid_thr)
nbpix_tot = np.sum(bpix_map)
nbpix_tbc = nbpix_tot - np.sum(bpix_map[circl_new])
bpix_map[circl_new] = 0
bpix_map_cumul = bpix_map_cumul + bpix_map
if verbose:
print('All bad pixels are corrected.')
if half_res_y:
frame = obj_tmp.copy()
frame_bpix = bpix_map_cumul.copy()
n_y = 2 * n_y
obj_tmp = np.zeros([n_y, n_x])
bpix_map_cumul = np.zeros([n_y, n_x])
for yy in range(n_y):
obj_tmp[yy] = frame[int(yy / 2)]
bpix_map_cumul[yy] = frame_bpix[int(yy / 2)]
return obj_tmp, bpix_map_cumul
def test_ellipse_generic():
img = np.zeros((4, 4), 'uint8')
rr, cc = ellipse(1.5, 1.5, 1.1, 1.7)
img[rr, cc] = 1
img_ = np.array([
[0, 0, 0, 0],
[1, 1, 1, 1],
[1, 1, 1, 1],
[0, 0, 0, 0],
])
assert_array_equal(img, img_)
img = np.zeros((5, 5), 'uint8')
rr, cc = ellipse(2, 2, 1.7, 1.7)
img[rr, cc] = 1
img_ = np.array([
[0, 0, 0, 0, 0],
[0, 1, 1, 1, 0],
[0, 1, 1, 1, 0],
[0, 1, 1, 1, 0],
[0, 0, 0, 0, 0],
])
assert_array_equal(img, img_)
img = np.zeros((10, 10), 'uint8')
rr, cc = ellipse(5, 5, 3, 4)
img[rr, cc] = 1
img_ = np.array([
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 1, 1, 1, 1, 1, 0, 0],
[0, 0, 1, 1, 1, 1, 1, 1, 1, 0],
[0, 0, 1, 1, 1, 1, 1, 1, 1, 0],
[0, 0, 1, 1, 1, 1, 1, 1, 1, 0],
[0, 0, 0, 1, 1, 1, 1, 1, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
])
assert_array_equal(img, img_)
img = np.zeros((10, 10), 'uint8')
rr, cc = ellipse(4.5, 5, 3.5, 4)
img[rr, cc] = 1
img_ = np.array([
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 1, 1, 1, 1, 1, 0, 0],
[0, 0, 1, 1, 1, 1, 1, 1, 1, 0],
[0, 0, 1, 1, 1, 1, 1, 1, 1, 0],
[0, 0, 1, 1, 1, 1, 1, 1, 1, 0],
[0, 0, 1, 1, 1, 1, 1, 1, 1, 0],
[0, 0, 0, 1, 1, 1, 1, 1, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
])
assert_array_equal(img, img_)
img = np.zeros((15, 15), 'uint8')
rr, cc = ellipse(7, 7, 3, 7)
img[rr, cc] = 1
img_ = np.array([
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0],
[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0],
[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0],
[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0],
[0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
])
assert_array_equal(img, img_)
poly = np.array((
(300, 300),
(480, 320),
(380, 430),
(220, 590),
(300, 300),
))
rr, cc = polygon(poly[:,0], poly[:,1], img.shape)
img[rr,cc,1] = 255
# fill circle
rr, cc = circle(200, 200, 100, img.shape)
img[rr,cc,:] = (255, 255, 0)
# fill ellipse
rr, cc = ellipse(300, 300, 100, 200, img.shape)
img[rr,cc,2] = 255
# circle
rr, cc = circle_perimeter(120, 400, 15)
img[rr, cc, :] = (255, 0, 0)
# ellipses
rr, cc = ellipse_perimeter(120, 400, 60, 20, orientation=math.pi / 4.)
img[rr, cc, :] = (255, 0, 255)
rr, cc = ellipse_perimeter(120, 400, 60, 20, orientation=-math.pi / 4.)
img[rr, cc, :] = (0, 0, 255)
rr, cc = ellipse_perimeter(120, 400, 60, 20, orientation=math.pi / 2.)
img[rr, cc, :] = (255, 255, 255)
plt.imshow(img)
# approximate subdivided polygon with Douglas-Peucker algorithm
appr_hand = approximate_polygon(new_hand, tolerance=0.02)
print("Number of coordinates:", len(hand), len(new_hand), len(appr_hand))
fig, (ax1, ax2) = plt.subplots(ncols=2, figsize=(9, 4))
ax1.plot(hand[:, 0], hand[:, 1])
ax1.plot(new_hand[:, 0], new_hand[:, 1])
ax1.plot(appr_hand[:, 0], appr_hand[:, 1])
# create two ellipses in image
img = np.zeros((800, 800), 'int32')
rr, cc = ellipse(250, 250, 180, 230, img.shape)
img[rr, cc] = 1
rr, cc = ellipse(600, 600, 150, 90, img.shape)
img[rr, cc] = 1
plt.gray()
ax2.imshow(img)
# approximate / simplify coordinates of the two ellipses
for contour in find_contours(img, 0):
coords = approximate_polygon(contour, tolerance=2.5)
ax2.plot(coords[:, 1], coords[:, 0], '-r', linewidth=2)
coords2 = approximate_polygon(contour, tolerance=39.5)
ax2.plot(coords2[:, 1], coords2[:, 0], '-g', linewidth=2)
print("Number of coordinates:", len(contour), len(coords), len(coords2))
def process_one(kdata, kdata_ref, acs_out, debug_path='debug_fig_sq'):
# Transpose to get the channel index in position 0, and rotate the image (if necessary)
#img_coils = np.rot90(np.transpose(img_coils, (2,0,1)),k=-1,axes=(1,2))
#print (img_coils.shape)
# Crop the image (if necessary)
#img_coils = img_coils[:,:,:]
# Get the dimensions
kspace_fully_sampled = kdata
[num_chan, N1, N2, Nt] = kspace_fully_sampled.shape
#mosaic(kspace_fully_sampled[:,:,:], 4, 8, 1, [0,2e-5], fig_title='kspace_fully_sampled', fig_size=(15,15), num_rot=0)
# Get the rssq image
'''
kspace_fully_sampled = np.fft.fftshift(np.fft.ifft(np.fft.ifftshift(kspace_fully_sampled, axes=3), axis=3), axes=3)
img_coils = np.fft.fftshift(np.fft.ifft2(np.fft.ifftshift(kspace_fully_sampled, axes=(1,2)), axes=(1,2)), axes=(1,2))
image_original0 = np.sqrt(np.sum(np.square(np.abs(img_coils)),axis=0))
for i in range(75,90):
image_original = image_original0[:,:,i]
mosaic(image_original, 1, 1, 1, [0,150], fig_title=f'image_original_{i}', fig_base_path=debug_path, fig_size=(10,10), num_rot=0)
raise
'''
#num_acs = (24, 24, 256)
num_acs = (22, 22, 256)
#num_acs = (N1, 30, 40)
Rx = 4
Ry = 2
acs_x, acs_y, acs_t = num_acs
acs_start_index_x = N1 // 2 - num_acs[0] // 2 # inclusive
acs_start_index_y = N2 // 2 - num_acs[1] // 2 # inclusive
acs_end_index_x = np.int(np.ceil(N1 / 2)) + num_acs[0] // 2 # exclusive
acs_end_index_y = np.int(np.ceil(N2 / 2)) + num_acs[1] // 2 # exclusive
acs_start_index_t = Nt // 2 - num_acs[2] // 2 # inclusive
acs_end_index_t = np.int(np.ceil(Nt / 2)) + num_acs[2] // 2 # exclusive
#print (acs_start_index_x, acs_end_index_x)
#print (acs_start_index_y, acs_end_index_y)
#print (acs_start_index_t, acs_end_index_t)
#kspace_undersampled_zero_filled = np.zeros(kspace_fully_sampled.shape, dtype=kspace_fully_sampled.dtype)
#kspace_undersampled_zero_filled[:,0:N1:Ry,0:N2:Rx,:] = kspace_fully_sampled[:,0:N1:Ry,0:N2:Rx,:]
#print (kspace_undersampled_zero_filled.shape)
print(kspace_fully_sampled.shape)
c, w, h, z = kspace_fully_sampled.shape
mask_ellip = np.zeros((w, h), dtype=np.uint8)
rr, cc = ellipse(w // 2, h // 2, w // 2, h // 2)
mask_ellip[rr, cc] = 1
print(mask_ellip.shape)
mask_ellip = np.expand_dims(mask_ellip, axis=0)
mask_ellip = np.repeat(mask_ellip, c, axis=0)
mask_ellip = np.expand_dims(mask_ellip, axis=-1)
mask_ellip = np.repeat(mask_ellip, z, axis=-1)
print(mask_ellip.shape)
mask = np.zeros(kspace_fully_sampled.shape)
mask[0:32, 0:N1:7, 0:N2:7, :] = 1
mask[0:32, 1:N1:7, 3:N2:7, :] = 1
mask[0:32, 2:N1:7, 6:N2:7, :] = 1
mask[0:32, 3:N1:7, 2:N2:7, :] = 1
mask[0:32, 4:N1:7, 5:N2:7, :] = 1
mask[0:32, 5:N1:7, 1:N2:7, :] = 1
mask[0:32, 6:N1:7, 4:N2:7, :] = 1
mask[32:64, 1:N1:7, 1:N2:7, :] = 1
mask[32:64, 2:N1:7, 4:N2:7, :] = 1
mask[32:64, 3:N1:7, 0:N2:7, :] = 1
mask[32:64, 4:N1:7, 3:N2:7, :] = 1
mask[32:64, 5:N1:7, 6:N2:7, :] = 1
mask[32:64, 6:N1:7, 2:N2:7, :] = 1
mask[32:64, 0:N1:7, 5:N2:7, :] = 1
mask[64:96, 2:N1:7, 2:N2:7, :] = 1
mask[64:96, 3:N1:7, 5:N2:7, :] = 1
mask[64:96, 4:N1:7, 1:N2:7, :] = 1
mask[64:96, 5:N1:7, 4:N2:7, :] = 1
mask[64:96, 6:N1:7, 0:N2:7, :] = 1
mask[64:96, 0:N1:7, 3:N2:7, :] = 1
mask[64:96, 1:N1:7, 6:N2:7, :] = 1
'''
mask[:, 0:N1:7, 0:N2:7, :] = 1
mask[:, 1:N1:7, 3:N2:7, :] = 1
mask[:, 2:N1:7, 6:N2:7, :] = 1
mask[:, 3:N1:7, 2:N2:7, :] = 1
mask[:, 4:N1:7, 5:N2:7, :] = 1
mask[:, 5:N1:7, 1:N2:7, :] = 1
mask[:, 6:N1:7, 4:N2:7, :] = 1
#print (mask[10,0:16,0:16,10])
#raise
'''
'''
mask[0:16, 0:N1:7, 0:N2:7, :] = 1
mask[0:16, 1:N1:7, 3:N2:7, :] = 1
mask[0:16, 2:N1:7, 6:N2:7, :] = 1
mask[0:16, 3:N1:7, 2:N2:7, :] = 1
mask[0:16, 4:N1:7, 5:N2:7, :] = 1
mask[0:16, 5:N1:7, 1:N2:7, :] = 1
mask[0:16, 6:N1:7, 4:N2:7, :] = 1
mask[16:32, 1:N1:7, 1:N2:7, :] = 1
mask[16:32, 2:N1:7, 4:N2:7, :] = 1
mask[16:32, 3:N1:7, 0:N2:7, :] = 1
mask[16:32, 4:N1:7, 3:N2:7, :] = 1
mask[16:32, 5:N1:7, 6:N2:7, :] = 1
mask[16:32, 6:N1:7, 2:N2:7, :] = 1
mask[16:32, 0:N1:7, 5:N2:7, :] = 1
mask[32:48, 2:N1:7, 2:N2:7, :] = 1
mask[32:48, 3:N1:7, 5:N2:7, :] = 1
mask[32:48, 4:N1:7, 1:N2:7, :] = 1
mask[32:48, 5:N1:7, 4:N2:7, :] = 1
mask[32:48, 6:N1:7, 0:N2:7, :] = 1
mask[32:48, 0:N1:7, 3:N2:7, :] = 1
mask[32:48, 1:N1:7, 6:N2:7, :] = 1
'''
#print (mask.shape)
#raise
kspace_undersampled_zero_filled0 = kspace_fully_sampled * mask * mask_ellip
#kspace_undersampled_zero_filled = kspace_undersampled_zero_filled0[0:16,...] + kspace_undersampled_zero_filled0[16:32,...] + kspace_undersampled_zero_filled0[32:48,...]
kspace_undersampled_zero_filled = kspace_undersampled_zero_filled0[
0:32, ...] + kspace_undersampled_zero_filled0[32:64, ...]
kspace_acs_zero_filled = np.zeros(kspace_fully_sampled.shape,
dtype=kspace_fully_sampled.dtype)
kspace_acs_zero_filled[:, acs_start_index_x:acs_end_index_x,
acs_start_index_y:acs_end_index_y,
acs_start_index_t:
acs_end_index_t] = kspace_fully_sampled[:,
acs_start_index_x:
acs_end_index_x,
acs_start_index_y:
acs_end_index_y,
acs_start_index_t:
acs_end_index_t]
print(kspace_acs_zero_filled.shape)
kspace_acs_cropped = np.zeros(
[num_chan, num_acs[0], num_acs[1], num_acs[2]],
dtype=kspace_fully_sampled.dtype)
kspace_acs_cropped[:, :, :, :] = kspace_fully_sampled[:, acs_start_index_x:
acs_end_index_x,
acs_start_index_y:
acs_end_index_y,
acs_start_index_t:
acs_end_index_t]
#tmp = kdata_ref[acs_start_index_x:acs_end_index_x,acs_start_index_y:acs_end_index_y, acs_start_index_t:acs_end_index_t]
#print (tmp.shape)
#print (acs_out.shape)
#print (tmp[:,12,100])
#print (acs_out[0,:,12,100])
#raise
#kspace_coils_acs_out = acs_out
kspace_coils_acs_out = acs_out[:, 1:-1, 1:-1, :]
#kspace_coils_acs_out = acs_out[:,acs_start_index_x:acs_end_index_x,acs_start_index_y:acs_end_index_y, acs_start_index_t:acs_end_index_t]
#kspace_coils_acs_out = kspace_acs_cropped
#kspace_coils_acs_out = np.expand_dims(kdata_ref[acs_start_index_x:acs_end_index_x,acs_start_index_y:acs_end_index_y, acs_start_index_t:acs_end_index_t], axis=0)
'''
tmp = np.expand_dims(kdata_ref[acs_start_index_x:acs_end_index_x,acs_start_index_y:acs_end_index_y, acs_start_index_t:acs_end_index_t], axis=0)
print (kspace_coils_acs_out.shape)
print (tmp.shape)
print (kspace_coils_acs_out[0,10:12,10:12,5])
print (tmp[0,10:12,10:12,5])
raise
'''
#print (kspace_coils_acs_out.shape)
#img_test = np.fft.fftshift(np.fft.ifft2(np.fft.ifftshift(kspace_coils_acs_out[0,:,:,0], axes=(0,1)), axes=(0,1)), axes=(0,1))
#mosaic(img_test, 1, 1, 1, [0,2e-5], fig_title='img_test', fig_base_path='debug_test', fig_size=(15,15), num_rot=0)
#kdata_ref[acs_start_index_x:acs_end_index_x,acs_start_index_y:acs_end_index_y, acs_start_index_t:acs_end_index_t] = acs_out[0,...] #put acs back to ref
img_coils_ref = np.fft.fftshift(np.fft.ifftn(np.fft.ifftshift(kdata_ref,
axes=(0, 1,
2)),
axes=(0, 1, 2)),
axes=(0, 1, 2))
#img_coils_ref = np.fft.fftshift(np.fft.ifft2(np.fft.ifftshift(kdata_ref, axes=(0,1)), axes=(0,1)), axes=(0,1))
image_original_ref = np.abs(img_coils_ref)
#image_original_ref = np.sqrt(np.sum(np.square(np.abs(img_coils_ref)),axis=0))
print('image_original_ref shape: ', image_original_ref.shape)
start_t = time.time()
name_image = f'{debug_path}/recon.mat'
name_weight = f'{debug_path}/weight.mat'
raki(num_chan,
N1,
N2,
Nt,
acs_start_index_x,
acs_end_index_x,
acs_start_index_y,
acs_end_index_y,
acs_start_index_t,
acs_end_index_t,
kspace_undersampled_zero_filled,
kspace_acs_zero_filled,
kspace_coils_acs_out,
kspace_acs_cropped,
Rx,
Ry,
debug_path,
name_image,
name_weight,
train_flag=True)
print(time.time() - start_t)
#show_raki_results(image_original_ref, debug_path, name_image)
show_raki_results_3d(image_original_ref, debug_path, name_image)
