def test_plot_plugin():
viewer = ImageViewer(data.moon())
plugin = PlotPlugin(image_filter=lambda x: x)
viewer += plugin
assert_equal(viewer.image, data.moon())
plugin._update_original_image(data.coins())
assert_equal(viewer.image, data.coins())
viewer.close()
def test_canny():
image = data.camera()
viewer = ImageViewer(image)
c = CannyPlugin()
viewer += c
canny_edges = viewer.show(False)
viewer.close()
edges = canny_edges[0][0]
assert edges.sum() == 2852
def show_disparity_map(disparity_map):
#Scale to zero:
image = np.subtract(disparity_map, np.min(disparity_map))
if np.max(image) != 0:
#Normalize and invert:
image = np.multiply(image, 255.0/float(np.max(image)))
#Show the stuff:
viewer = ImageViewer(image.astype(np.uint8))
viewer.show()
def test_measure():
image = data.camera()
viewer = ImageViewer(image)
m = Measure()
viewer += m
m.line_changed([(0, 0), (10, 10)])
assert_equal(str(m._length.text), '14.1')
assert_equal(str(m._angle.text[:5]), '135.0')
def test_color_histogram():
image = util.img_as_float(data.colorwheel())
viewer = ImageViewer(image)
ch = ColorHistogram(dock='right')
viewer += ch
assert_almost_equal(viewer.image.std(), 0.352, 3),
ch.ab_selected((0, 100, 0, 100)),
assert_almost_equal(viewer.image.std(), 0.325, 3)
def test_viewer_with_overlay():
img = data.coins()
ov = OverlayPlugin(image_filter=sobel)
viewer = ImageViewer(img)
viewer += ov
import tempfile
_, filename = tempfile.mkstemp(suffix='.png')
ov.color = 3
assert_equal(ov.color, 'yellow')
viewer.save_to_file(filename)
ov.display_filtered_image(img)
assert_equal(ov.overlay, img)
ov.overlay = None
assert_equal(ov.overlay, None)
ov.overlay = img
assert_equal(ov.overlay, img)
assert_equal(ov.filtered_image, img)
def test_viewer_with_overlay():
img = data.coins()
ov = OverlayPlugin(image_filter=sobel)
viewer = ImageViewer(img)
viewer += ov
import tempfile
_, filename = tempfile.mkstemp(suffix='.png')
ov.color = 3
assert_equal(ov.color, 'yellow')
viewer.save_to_file(filename)
ov.display_filtered_image(img)
assert_equal(ov.overlay, img)
ov.overlay = None
assert_equal(ov.overlay, None)
ov.overlay = img
assert_equal(ov.overlay, img)
assert_equal(ov.filtered_image, img)
def show_best_match(image_id, codebook, metric):
bows = [x[1] for x in codebook]
if metric == 'euclidean':
m = euclidean_distance
elif metric == 'kullback-leibler':
m = symmetric_kullback_leibler
elif metric == 'bhattacharyya':
m = bhattacharyya_distance
elif metric == 'common words':
m = common_words
else:
return "Error"
best_match_path = codebook[find_best_match(image_id, bows, m)][0]
img = imread(best_match_path)
view = ImageViewer(img)
view.show()
def main():
imagefile = io.imread("colors.jpg")
recolored_image = copy.deepcopy(imagefile)
grayscale_matrix = np.array([0.2125, 0.7154, 0.0721])
# loop through every pixel
for pixel_row_index in range(len(imagefile)):
for pixel_index in range(len(imagefile[pixel_row_index])):
if (imagefile[pixel_row_index][pixel_index][0] > 165
and imagefile[pixel_row_index][pixel_index][1] < 150
and imagefile[pixel_row_index][pixel_index][2] < 150) and (
imagefile[pixel_row_index][pixel_index][1] <
imagefile[pixel_row_index][pixel_index][2] + 15):
recolored_image[pixel_row_index][pixel_index] = imagefile[
pixel_row_index][pixel_index]
else:
recolored_image[pixel_row_index][pixel_index] = imagefile[
pixel_row_index][pixel_index].dot(grayscale_matrix)
# transform rgb to hsv for hue
color.rgb2hsv(imagefile)
hue_original = get_hues(imagefile)
color.rgb2hsv(recolored_image)
hue_recolored = get_hues(recolored_image)
#create plots
fig, axs = plt.subplots(2)
num_bins = 36
n, bins, patches = axs[0].hist(hue_original,
num_bins,
facecolor='blue',
alpha=0.5)
n, bins, patches = axs[1].hist(hue_recolored,
num_bins,
facecolor='red',
alpha=0.5)
plt.show()
#grayscale = color.rgb2gray(imagefile)
viewer = ImageViewer(recolored_image)
viewer.show()
def preserve_color_cielch(content, stylized, image_name):
# extract info and convert to CIE-LCh for each image
rgbContent = io.imread(content)
labContent = color.lab2lch(
color.xyz2lab(color.rgb2xyz(numpy.asarray(rgbContent))))
labContentArray = numpy.array(labContent)
rgbStyle = io.imread(stylized)
labStyle = color.lab2lch(
color.xyz2lab(color.rgb2xyz(numpy.asarray(rgbStyle))))
labStyleArray = numpy.array(labStyle)
# color transfer
for i in range(len(labContentArray)):
for j in range(len(labContentArray[0])):
labContentArray[i][j][0] = labStyleArray[i][j][0]
labContentArray = color.xyz2rgb(
color.lab2xyz(color.lch2lab(labContentArray)))
viewer = ImageViewer(labContentArray)
viewer.show()
def remove_object_loop(img):
print("Remove object")
viewer = ImageViewer(img)
def on_enter(extens):
coord = np.int64(extens)
[x0, y0] = [coord[2], coord[0]]
[x1, y1] = [coord[3], coord[1]]
print([x0, y0], [x1, y1])
(x, y) = (x0, y0)
(len_x, len_y) = (x1 - x, y1 - y)
viewer.image = remove_object(viewer.image, x0, y0, len_x, len_y)
rect_tool = RectangleTool(viewer, on_enter=on_enter)
viewer.show()
return viewer.image
def test_line_profile_dynamic():
"""Test a line profile updating after an image transform"""
image = data.coins()[:-50, :] # shave some off to make the line lower
image = skimage.img_as_float(image)
viewer = ImageViewer(image)
lp = LineProfile(limits='dtype')
viewer += lp
line = lp.get_profiles()[-1][0]
assert line.size == 129
assert_almost_equal(np.std(viewer.image), 0.208, 3)
assert_almost_equal(np.std(line), 0.229, 3)
assert_almost_equal(np.max(line) - np.min(line), 0.725, 1)
viewer.image = skimage.img_as_float(median(image, selem=disk(radius=3)))
line = lp.get_profiles()[-1][0]
assert_almost_equal(np.std(viewer.image), 0.198, 3)
assert_almost_equal(np.std(line), 0.220, 3)
assert_almost_equal(np.max(line) - np.min(line), 0.639, 1)
def main():
filename_queue = tf.train.string_input_producer([tfrecords_filename])
image = read_and_decode(filename_queue)
init_op = tf.initialize_all_variables()
with tf.Session() as sess:
sess.run(init_op)
coord = tf.train.Coordinator()
threads = tf.train.start_queue_runners(coord=coord)
for i in range(3):
img = sess.run([image])
img = img[0]
print(img[0])
viewer = ImageViewer(img)
viewer.show()
coord.request_stop()
coord.join(threads)
def go(img):
def preprocess(img):
U = {}
V = {}
s = {}
for i in (0, 1, 2):
res = svd(img[..., i], full_matrices=False)
U[i] = np.mat(res[0])
V[i] = np.mat(res[2])
s[i] = res[1]
return U, V, s
U, V, s = preprocess(img)
def display(_, R, G, B):
rgb = (R, G, B)
res = np.zeros_like(img)
for i in (0, 1, 2):
_s = s[i].copy()
_s[rgb[i]:] = 0
S = np.mat(np.diag(_s))
res[..., i] = U[i] * S * V[i]
return res
def display_all(_, RGB):
return display(_, RGB, RGB, RGB)
viewer = ImageViewer(img)
plugin = Plugin(image_filter=display)
plugin.name = ""
plugin += Slider('R', 0, len(s[0]), len(s[0]), 'int')
plugin += Slider('G', 0, len(s[0]), len(s[1]), 'int')
plugin += Slider('B', 0, len(s[0]), len(s[2]), 'int')
viewer += plugin
plugin = Plugin(image_filter=display_all)
plugin.name = ""
plugin += Slider('RGB', 0, len(s[0]), len(s[0]), 'int')
viewer += plugin
viewer.show()
def test_line_profile_dynamic():
"""Test a line profile updating after an image transform"""
image = data.coins()[:-50, :] # shave some off to make the line lower
image = skimage.img_as_float(image)
viewer = ImageViewer(image)
lp = LineProfile(limits='dtype')
viewer += lp
line = lp.get_profiles()[-1][0]
assert line.size == 129
assert_almost_equal(np.std(viewer.image), 0.208, 3)
assert_almost_equal(np.std(line), 0.229, 3)
assert_almost_equal(np.max(line) - np.min(line), 0.725, 1)
viewer.image = skimage.img_as_float(median(image,
selem=disk(radius=3)))
line = lp.get_profiles()[-1][0]
assert_almost_equal(np.std(viewer.image), 0.198, 3)
assert_almost_equal(np.std(line), 0.220, 3)
assert_almost_equal(np.max(line) - np.min(line), 0.639, 1)
def get_crop(image):
'''
Show the original image to allow the user to crop any extraneous
information out of the frame.
:param image: 2D numpy array grayscale image
:return: list of [left, right, top, bottom] values for the edges of the
bounding box
'''
plt.ioff()
print('Waiting for input, please crop the image as desired and hit enter')
viewer = ImageViewer(image)
rect_tool = RectangleTool(viewer, on_enter=viewer.closeEvent)
viewer.show()
bounds = np.array(rect_tool.extents)
if (bounds[0] < 0 or bounds[1] > image.shape[1] or bounds[2] < 0
or bounds[3] > image.shape[0]):
print(f'Bounds = {bounds} and shape = {image.shape}')
raise ValueError('Bounds outside image, make sure to select within')
plt.ion()
return np.array(np.round(bounds), dtype=int)
def test_label_painter():
image = data.camera()
moon = data.moon()
viewer = ImageViewer(image)
lp = LabelPainter()
viewer += lp
assert_equal(lp.radius, 5)
lp.label = 1
assert_equal(str(lp.label), '1')
lp.label = 2
assert_equal(str(lp.paint_tool.label), '2')
assert_equal(lp.paint_tool.radius, 5)
lp._on_new_image(moon)
assert_equal(lp.paint_tool.shape, moon.shape)
def test_base_tool():
img = data.moon()
viewer = ImageViewer(img)
tool = CanvasToolBase(viewer)
tool.set_visible(False)
tool.set_visible(True)
do_event(viewer, 'key_press', key='enter')
tool.redraw()
tool.remove()
tool = CanvasToolBase(viewer, useblit=False)
tool.redraw()
def test_base_tool():
img = data.moon()
viewer = ImageViewer(img)
tool = CanvasToolBase(viewer.ax)
tool.set_visible(False)
tool.set_visible(True)
enter = create_mouse_event(viewer.ax, key='enter')
tool._on_key_press(enter)
tool.redraw()
tool.remove()
tool = CanvasToolBase(viewer.ax, useblit=False)
tool.redraw()
def viewer():
image = skimage.data.coins()
viewer = ImageViewer(image)
viewer.show()
viewer += LineProfile(viewer)
overlay, data = viewer.show()[0]
#--------------------
denoise_plugin = Plugin(
image_filter=skimage.restoration.denoise_tv_bregman)
denoise_plugin += Slider('weight', 0.01, 0.5, update_on='release')
denoise_plugin += SaveButtons()
viewer = ImageViewer(image)
viewer += denoise_plugin
denoised = viewer.show()[0][0]
def test_thick_line_tool():
img = data.camera()
viewer = ImageViewer(img)
tool = ThickLineTool(viewer, maxdist=10)
tool.end_points = get_end_points(img)
do_event(viewer, 'scroll', button='up')
assert_equal(tool.linewidth, 2)
do_event(viewer, 'scroll', button='down')
assert_equal(tool.linewidth, 1)
do_event(viewer, 'key_press', key='+')
assert_equal(tool.linewidth, 2)
do_event(viewer, 'key_press', key='-')
assert_equal(tool.linewidth, 1)
def test_thick_line_tool():
img = data.camera()
viewer = ImageViewer(img)
tool = ThickLineTool(viewer, maxdist=10, line_props=dict(color='red'),
handle_props=dict(markersize=5))
tool.end_points = get_end_points(img)
do_event(viewer, 'scroll', button='up')
assert_equal(tool.linewidth, 2)
do_event(viewer, 'scroll', button='down')
assert_equal(tool.linewidth, 1)
do_event(viewer, 'key_press', key='+')
assert_equal(tool.linewidth, 2)
do_event(viewer, 'key_press', key='-')
assert_equal(tool.linewidth, 1)
def show_image_sample(image_directory,
grid_shape,
extensions=['jpg'],
show=True,
output_filename=None,
thumbnail_size=100,
assert_enough_images=True):
"""Creates image of a grid of images sampled from `image_directory`."""
if isinstance(thumbnail_size, int):
size = (thumbnail_size, thumbnail_size)
else:
size = thumbnail_size
def is_image(filename):
return any(filename.lower().endswith(x.lower()) for x in extensions)
image_files = [
os.path.join(image_directory, fn) for fn in os.listdir(image_directory)
if is_image(fn)
]
assert len(grid_shape) == 2
if assert_enough_images:
assert grid_shape[0] * grid_shape[1] <= len(image_files)
sample = np.random.choice(image_files, grid_shape)
grid = []
for i in range(grid_shape[0]):
row = []
for j in range(grid_shape[1]):
row.append(resize(imread(sample[i, j]), size))
grid.append(np.concatenate(row, axis=1))
grid = (255 * np.concatenate(grid)).astype('uint8')
if show:
ImageViewer(grid).show()
if output_filename is not None:
imwrite(output_filename, grid)
def test_line_tool():
img = data.camera()
viewer = ImageViewer(img)
tool = LineTool(viewer, maxdist=10)
tool.end_points = get_end_points(img)
assert_equal(tool.end_points, np.array([[170, 256], [341, 256]]))
# grab a handle and move it
do_event(viewer, 'mouse_press', xdata=170, ydata=256)
do_event(viewer, 'move', xdata=180, ydata=260)
do_event(viewer, 'mouse_release')
assert_equal(tool.geometry, np.array([[180, 260], [341, 256]]))
# create a new line
do_event(viewer, 'mouse_press', xdata=10, ydata=10)
do_event(viewer, 'move', xdata=100, ydata=100)
do_event(viewer, 'mouse_release')
assert_equal(tool.geometry, np.array([[100, 100], [10, 10]]))
def test_thick_line_tool():
img = data.camera()
viewer = ImageViewer(img)
tool = ThickLineTool(viewer.ax, maxdist=10)
tool.end_points = get_end_points(img)
scroll_up = create_mouse_event(viewer.ax, button='up')
tool.on_scroll(scroll_up)
assert_equal(tool.linewidth, 2)
scroll_down = create_mouse_event(viewer.ax, button='down')
tool.on_scroll(scroll_down)
assert_equal(tool.linewidth, 1)
key_up = create_mouse_event(viewer.ax, key='+')
tool.on_key_press(key_up)
assert_equal(tool.linewidth, 2)
key_down = create_mouse_event(viewer.ax, key='-')
tool.on_key_press(key_down)
assert_equal(tool.linewidth, 1)
def test_line_tool():
img = data.camera()
viewer = ImageViewer(img)
tool = LineTool(viewer.ax, maxdist=10)
tool.end_points = get_end_points(img)
assert_equal(tool.end_points, np.array([[170, 256], [341, 256]]))
# grab a handle and move it
grab = create_mouse_event(viewer.ax, xdata=170, ydata=256)
tool.on_mouse_press(grab)
move = create_mouse_event(viewer.ax, xdata=180, ydata=260)
tool.on_move(move)
tool.on_mouse_release(move)
assert_equal(tool.geometry, np.array([[180, 260], [341, 256]]))
# create a new line
new = create_mouse_event(viewer.ax, xdata=10, ydata=10)
tool.on_mouse_press(new)
move = create_mouse_event(viewer.ax, xdata=100, ydata=100)
tool.on_move(move)
tool.on_mouse_release(move)
assert_equal(tool.geometry, np.array([[100, 100], [10, 10]]))
def test_plugin():
img = skimage.img_as_float(data.moon())
viewer = ImageViewer(img)
def median_filter(img, radius=3):
return median(img, selem=disk(radius=radius))
plugin = Plugin(image_filter=median_filter)
viewer += plugin
plugin += Slider('radius', 1, 5)
assert_almost_equal(np.std(viewer.image), 12.556, 3)
plugin.filter_image()
assert_almost_equal(np.std(viewer.image), 12.931, 3)
plugin.show()
plugin.close()
plugin.clean_up()
img, _ = plugin.output()
assert_equal(img, viewer.image)
def test_rect_tool():
img = data.camera()
viewer = ImageViewer(img)
tool = RectangleTool(viewer, maxdist=10)
tool.extents = (100, 150, 100, 150)
assert_equal(tool.corners, ((100, 150, 150, 100), (100, 100, 150, 150)))
assert_equal(tool.extents, (100, 150, 100, 150))
assert_equal(tool.edge_centers,
((100, 125.0, 150, 125.0), (125.0, 100, 125.0, 150)))
assert_equal(tool.geometry, (100, 150, 100, 150))
# grab a corner and move it
do_event(viewer, 'mouse_press', xdata=100, ydata=100)
do_event(viewer, 'move', xdata=120, ydata=120)
do_event(viewer, 'mouse_release')
assert_equal(tool.geometry, [120, 150, 120, 150])
# create a new line
do_event(viewer, 'mouse_press', xdata=10, ydata=10)
do_event(viewer, 'move', xdata=100, ydata=100)
do_event(viewer, 'mouse_release')
assert_equal(tool.geometry, [10, 100, 10, 100])
def show_matching(img, img2, matching):
print "matching..."
ip_match = build_match_dic(img, img2, matching)
print "Constructing intermediate image..."
padding = 5 #padding around the edges
bar = np.ndarray((img.shape[0], 5))
bar.fill(1.0)
viewer = ImageViewer(img)
viewer.show()
img3 = np.column_stack((img, bar, img2))
viewer = ImageViewer(img3)
viewer.show()
img3 = img_as_ubyte(img3)
viewer = ImageViewer(img3)
viewer.show()
img3 = np.pad(img3, pad_width=padding, mode='constant', constant_values=(0))
viewer = ImageViewer(img3)
viewer.show()
print "Drawing lines..."
colimg = color.gray2rgb(img3)
for k,v in random.sample(ip_match.items(), int(float(len(ip_match.keys()))*0.005)):
#Choose a random colour:
col = [random.randint(0,255),random.randint(0,255),random.randint(0,255)]
#Calculate coordinates after padding:
x1 = k[0]+padding
y1 = k[1]+padding
x2 = v[0]+padding
y2 = v[1] + img.shape[1]+bar.shape[1]+padding
#Draw the points in both images:
rr, cc = circle_perimeter(x1, y1, 3)
colimg[rr, cc] = col
rr, cc = circle_perimeter(x2, y2, 3)
colimg[rr, cc] = col
#Draw a line between the points:
rr, cc = line(x1,y1,x2,y2)
colimg[rr, cc] = col
#Show the result:
viewer = ImageViewer(colimg)
viewer.show()
def check(train_set_x, train_set_y, val_set_x, val_set_y):
print(train_set_y[0])
viewer = ImageViewer(train_set_x[0])
viewer.show()
print(train_set_y[10])
viewer = ImageViewer(train_set_x[10])
viewer.show()
print(train_set_y[32])
viewer = ImageViewer(train_set_x[32])
viewer.show()
print(train_set_y[56])
viewer = ImageViewer(train_set_x[56])
viewer.show()
print(train_set_y[76])
viewer = ImageViewer(train_set_x[76])
viewer.show()
print(train_set_y[119])
viewer = ImageViewer(train_set_x[119])
viewer.show()
print(val_set_y[5])
viewer = ImageViewer(val_set_x[5])
viewer.show()
print(val_set_y[13])
viewer = ImageViewer(val_set_x[13])
viewer.show()
print(val_set_y[32])
viewer = ImageViewer(val_set_x[32])
viewer.show()
print(val_set_y[51])
viewer = ImageViewer(val_set_x[51])
viewer.show()
print(val_set_y[75])
viewer = ImageViewer(val_set_x[75])
viewer.show()
print(val_set_y[145])
viewer = ImageViewer(val_set_x[145])
viewer.show()
R, G, B = 0, 1, 2 # select regions where red is less than 100 mask = source[R].point(lambda i: i < 100 and 255) # process the green band out = source[G].point(lambda i: i * 0.7) # paste the processed band back, but only where red was < 100 source[G].paste(out, None, mask) # build a new multiband image im = Image.merge(img.mode, source) # im.show() # Python Imaging Library 使用笛卡尔坐标系, 使用 (0,0) 表示左上角. 值得注意的是, 坐标点表示的是一个像素的左上角, 而表示像素的中央则是 (0.5,0.5) # skimage 部分 camera = data.camera() print(camera, camera.shape, type(camera)) # 打开本地文件 dancing = io.imread(fp) print(dancing, dancing.shape, type(dancing)) # Images manipulated by scikit-image are simply NumPy arrays print(dancing.min(), dancing.max(), dancing.mean()) # plt.imshow(dancing) # plt.show() viewer = ImageViewer(dancing) viewer.show()
if self.active_handle is None:
# New rectangle
x1 = self.eventpress.xdata
y1 = self.eventpress.ydata
x2, y2 = event.xdata, event.ydata
else:
x1, x2, y1, y2 = self._extents_on_press
if self.active_handle in ['E', 'W'] + self._corner_order:
x2 = event.xdata
if self.active_handle in ['N', 'S'] + self._corner_order:
y2 = event.ydata
self.extents = (x1, x2, y1, y2)
self.callback_on_move(self.geometry)
@property
def geometry(self):
return self.extents
if __name__ == '__main__': # pragma: no cover
from skimage.viewer import ImageViewer
from skimage import data
viewer = ImageViewer(data.camera())
rect_tool = RectangleTool(viewer)
viewer.show()
print("Final selection:")
rect_tool.callback_on_enter(rect_tool.extents)
def test_viewer():
astro = data.astronaut()
coins = data.coins()
view = ImageViewer(astro)
import tempfile
_, filename = tempfile.mkstemp(suffix='.png')
view.show(False)
view.close()
view.save_to_file(filename)
view.open_file(filename)
assert_equal(view.image, astro)
view.image = coins
assert_equal(view.image, coins),
view.save_to_file(filename),
view.open_file(filename),
view.reset_image(),
assert_equal(view.image, coins)
import openslide
import numpy as np
import skimage
from skimage.viewer import ImageViewer
import PIL
import matplotlib.pyplot as plt
slide = openslide.open_slide(
"../CAMELYON/data/patient_013/patient_013_node_0.tif")
slide
level = 6
dims = slide.level_dimensions[level]
slide.level_downsamples[level]
pixelarray = np.array(slide.read_region((0, 0), level, dims))
pixelarray.shape
viewer = ImageViewer(pixelarray)
viewer.show()
from skimage import data
from skimage.filters.rank import median
from skimage.morphology import disk
from skimage.viewer import ImageViewer
from skimage.viewer.widgets import Slider, OKCancelButtons, SaveButtons
from skimage.viewer.plugins.base import Plugin
def median_filter(image, radius):
return median(image, selem=disk(radius))
image = data.coins()
viewer = ImageViewer(image)
plugin = Plugin(image_filter=median_filter)
plugin += Slider('radius', 2, 10, value_type='int')
plugin += SaveButtons()
plugin += OKCancelButtons()
viewer += plugin
viewer.show()
def test_viewer():
astro = data.astronaut()
coins = data.coins()
view = ImageViewer(astro)
import tempfile
_, filename = tempfile.mkstemp(suffix='.png')
view.show(False)
view.close()
view.save_to_file(filename)
view.open_file(filename)
assert_equal(view.image, astro)
view.image = coins
assert_equal(view.image, coins),
view.save_to_file(filename),
view.open_file(filename),
view.reset_image(),
assert_equal(view.image, coins)
from skimage import data from skimage.viewer import ImageViewer from skimage.viewer.plugins.lineprofile import LineProfile image = data.camera() viewer = ImageViewer(image) viewer += LineProfile() line, profile = viewer.show()[0]
from skimage import data
from skimage.filters import canny
from skimage.viewer import ImageViewer
from skimage.viewer.widgets import Slider
from skimage.viewer.widgets.history import SaveButtons
from skimage.viewer.plugins.overlayplugin import OverlayPlugin
image = data.camera()
# You can create a UI for a filter just by passing a filter function...
plugin = OverlayPlugin(image_filter=canny)
# ... and adding widgets to adjust parameter values.
plugin += Slider('sigma', 0, 5)
plugin += Slider('low threshold', 0, 255)
plugin += Slider('high threshold', 0, 255)
# ... and we can also add buttons to save the overlay:
plugin += SaveButtons(name='Save overlay to:')
# Finally, attach the plugin to an image viewer.
viewer = ImageViewer(image)
viewer += plugin
canny_edges = viewer.show()[0][0]
from skimage.viewer.plugins.canny import CannyPlugin
def line_image(shape, lines):
image = np.zeros(shape, dtype=bool)
for end_points in lines:
end_points = np.asarray(end_points)[:, ::-1]
image[draw.line(*np.ravel(end_points))] = 1
return image
def hough_lines(image, *args, **kwargs):
lines = probabilistic_hough_line(image, threshold=0.5, *args, **kwargs)
image = line_image(image.shape, lines)
return image
image = data.camera()
canny_viewer = ImageViewer(image)
canny_plugin = CannyPlugin()
canny_viewer += canny_plugin
hough_plugin = OverlayPlugin(image_filter=hough_lines)
hough_plugin += Slider('line length', 0, 100, value=100, update_on='move')
hough_plugin += Slider('line gap', 0, 20, value=0, update_on='move')
hough_viewer = ImageViewer(canny_plugin)
hough_viewer += hough_plugin
canny_viewer.show()
def show(self):
viewer = ImageViewer(self.image)
viewer.show()
def setup_line_profile(image, limits='image'):
viewer = ImageViewer(skimage.img_as_float(image))
plugin = LineProfile(limits=limits)
viewer += plugin
return plugin
from skimage import data from skimage.viewer import ImageViewer from skimage.viewer.plugins.measure import Measure image = data.camera() viewer = ImageViewer(image) viewer += Measure() viewer.show()
def viewImage(image):
viewer = ImageViewer(image)
viewer.show()
from skimage import data from skimage.viewer import ImageViewer from skimage.viewer.plugins.lineprofile import LineProfile image = data.chelsea() viewer = ImageViewer(image) viewer += LineProfile() line, rgb_profiles = viewer.show()[0]
def testinterp(noiselevel, pointfrac):
# Definitions
dimx = 100
dimy = 100
dens = 0.1
dx2 = dimx/2
dy2 = dimy/2
startval = 3.0
l = 0.01
maxitt = 10
# define ground truth
GT = np.zeros((dimx, dimy))
GT[0:dx2,0:dy2] = - 10.0
GT[0:dx2,dy2+1:] = 0.0
GT[dx2+1:,0:dy2] = 5.0
GT[dx2+1:,dy2+1:] = 15.0
# generate data
Data = GT + noiselevel*np.random.randn(dimx,dimy)
points = (np.random.rand(dimx,dimy) < pointfrac)
values = np.multiply((1*points), Data)
start = startval*np.ones((dimx,dimy))
print "Interpolating..."
Ipim = interp(start, points, values, maxitt, l)
print "Done!"
# compare and display result
err1 = (Data - Ipim)
err = np.absolute(Data - Ipim)
errsq = err1.dot(err1)
rms= math.sqrt(errsq.sum())/float(dimx*dimy)
print 'average reconstruction error:', rms
shGT = GT - GT.min()
shGT = shGT/float(shGT.max())
shIp = Ipim - Ipim.min()
shIp = shIp/float(shIp.max())
sher = err - err.min()
sher = sher/float(sher.max())
#h = figure(1);
#subplot(2,2,1);
print "ground truth"
view = ImageViewer(shGT)
view.show()
print 'Reconstruction'
view = ImageViewer(shIp)
view.show()
print 'Sparse point position'
view = ImageViewer(points)
view.show()
print 'Reconstruction error'
view = ImageViewer(sher)
view.show()
class SignalRetriever():
fun = lambda x, y, z: np.logical_or(x, np.logical_or(y, z))
view_func = lambda x: ImageViewer(x).show()
@staticmethod
def get_baseline(image_src):
pass
@staticmethod
def transform_to_binary_image(image_src, corrector=35):
#La funcion transforma la imagen a blanco y negro haciendo su mejor esfuerzo
#por preservar las líneas de la señal del ECG.
# Esta funcion realiza resalta los negros dado un determinado factor para
# visualizar la senial.
fun_corrector = lambda x: lambda y: int(
x == y) * y * corrector // 100 + int(x != y) * y
fun = SignalRetriever.fun
#Lectura de la imagen y posterior ajuste de Gamma (Para que eliminar difuminadso)
image = sexp.adjust_gamma(image_src, 1.25)
#Datos importantes para la imagen
try:
image_aux = np.reshape(
image, (image.shape[0] * image.shape[1], image.shape[2]))
except IndexError:
pass
#El color que mas aparece
color_dominance = np.sum(image_aux, 0)
color_max = np.max(color_dominance)
index_max = np.where(color_dominance == color_max)
index_max = index_max[0][0]
#Threshold para cada canal de la imagen
rgb_t = list(
map(lambda x: sfil.threshold_isodata(x),
[image_aux[:, i] for i in range(3)]))
fun_corrector = fun_corrector(rgb_t[index_max])
#Aplicando threshold a cada canal y unificando
image_aux = fun(image_aux[:, 0] > fun_corrector(rgb_t[0]),
image_aux[:, 1] > fun_corrector(rgb_t[1]),
image_aux[:, 2] > fun_corrector(rgb_t[2]))
#Redimensionando la imagen.
bn_im = np.reshape(image_aux, (image.shape[0], image.shape[1]))
return bn_im
@staticmethod
def retrieve_signal(points, image_src):
# Esta funcion toma los pixeles marcados y realiza una interpolacion con los pixeles
# obtenido del procesamiento de la imagen
y_standarized = (image_src.shape[0] - points[0]).astype(float)
inter_func = interp1d(points[1], y_standarized)
return inter_func
@staticmethod
def plot_digital_ecg(inter_func, points):
#Funcion de prueba para ver como se ve la señal luego de una interpolación.
x = np.linspace(np.min(points[1]), np.max(points[1]), 750)
plt.plot(x, inter_func(x))
plt.axes().set_aspect('equal', 'box')
plt.show()
@staticmethod
def sample_signal(inter_func, points):
# Guardando archivo de la señal.
x = np.linspace(np.min(points[1]), np.max(points[1]), 750)
x = inter_func(x)
sdx = np.std(x)
x -= np.mean(x)
x /= sdx
x = np.reshape(x, (x.shape[0], 1))
# record = wfdb.Record(recordname='Test1',fs=300,nsig=1,siglen=750,p_signals=x,
# filename=['test.dat'],baseline=[-1],units=['mV'],signame=['ECG'])
# wfdb.plotrec(record,title='Test')
return x
def __init__(self, *args, **kwargs):
self.images_src = args
self.dir = kwargs.get('dir')
exclude = kwargs.get('exclude')
# Hay archivos para excluir ?
if exclude:
#exclude_files = open(exclude,'r')
exclude_files = ['img_0.jpg', 'img_0_II_long.jpg']
# json parse
self.images_src = list(
map(
lambda x: self.dir + '/' + x,
filter(
lambda x: x not in exclude_files or x[-6:] == 'bn.jpg',
self.images_src)))
def get_multisignal_from_images(self):
images = sio.imread_collection(self.images_src)
i = 0
array_signal = []
for image in images:
points, bn_image = SignalRetriever.transform_to_binary_image(
image, corrector=15)
#imsave(self.dir+'/'+'img_{0}_bn.jpg'.format(i),bn_image)
inter_func = SignalRetriever.retrieve_signal(points, image)
x = SignalRetriever.sample_signal(inter_func, points)
# record = wfdb.Record(recordname='Test'+str(i),fs=300,nsig=1,siglen=750,p_signals=x,
# filename=['test.dat'],baseline=[50],units=['mV'],signame=['ECG'],adcgain=[200],fmt=['16'],checksum=[0],
# adcres=[16],adczero=[0],initvalue=[0],blocksize=[0], d_signals=None)
#array_signal = np.concatenate((array_signal,SignalRetriever.sample_signal(inter_func,points)))
array_signal.append(x)
i += 1
return array_signal
import skimage as sk from skimage.viewer import ImageViewer #import TomographyTools as tt #import TomographyTools.data_management as dm #import TomographyTools.reconstruciton as rn import TomographyTools.image_processing as ip inputDir = "/home/[email protected]/data-scratch/20171006_ChaiLor_PaintData/rec20171005_162920_TiO2_sample1/" inputDirs = ['/home/[email protected]/data-scratch/20171006_ChaiLor_PaintData/rec20171005_155036_TiO2_control_scan2', "/home/[email protected]/data-scratch/20171006_ChaiLor_PaintData/rec20171005_162920_TiO2_sample1/", '/home/[email protected]/data-scratch/20171006_ChaiLor_PaintData/rec20171005_165313_TiO2_sample2', '/home/[email protected]/data-scratch/20171006_ChaiLor_PaintData/rec20171005_172237_TiO2_sample3'] if True: ip.convert_DirectoryTo8Bit(inputpath=inputDirs[0],data_min=-5.0,data_max=7.0) ip.convert_DirectoryTo8Bit(inputpath=inputDirs[1],data_min=-5.0,data_max=7.0) ip.convert_DirectoryTo8Bit(inputpath=inputDirs[2],data_min=-5.0,data_max=7.0) ip.convert_DirectoryTo8Bit(inputpath=inputDirs[3],data_min=-5.0,data_max=7.0) if False: imageStack = ip.loadDataStack(inputDir) imageStack8Bit = ip.convert8bit(imageStack,-5.0,7.0) ip.saveTiffStack(imageStack,filename="testStack_8Bit") image = imageStack8Bit[:,:,50] viewer = ImageViewer(image); viewer.show()
