def measure_focused_roi(im, roi, area, focus_points, debug=False):
g = Grid(cv.GetSize(im))
canvas = image.new_from(im)
cv.Set(canvas, 0)
focus_in_roi = image.And(focus_points, roi)
if debug:
image.show(focus_in_roi, "ROI + Focused Points")
densities = []
points = convert_to_points(focus_in_roi)
groups = form_groups(points, estimated_size=24, iter=5)
for group in groups:
ch = ConvexHull(map(lambda x: (x[0], x[1]), group))
ppp = ch.points_per_pixel()
a = int(ppp * 255)
ch.draw_filled_hull(canvas, rgb=(a,a,a))
if debug:
image.show(canvas, "Focused Regions in ROI")
quadrants = g.split_in_four(canvas)
sums = []
for i,quad in enumerate(quadrants):
sums.append(cv.Sum(quad)[0] / float(area/4))
arr = array(sums)
print arr.mean(), arr.std()
diff = max(sums) - min(sums)
return diff, arr.std()
def draw_mask(img, cx, cy, r, out=0):
ss = np.shape(img)
x = np.arange(ss[0])
x = np.transpose(np.tile(x, [ss[1], 1]))
y = np.arange(ss[1])
y = np.tile(y, [ss[0], 1])
x = x - cx
y = y - cy
nr = np.size(r)
if nr == 2:
rr = x * x / r[0]**2 + y * y / r[1]**2
r1 = 1
else:
rr = x * x + y * y
r1 = r**2
pp = np.where(rr < r1)
img1 = img.copy()
if out == 1:
img1[pp] = 0
else:
img1[pp] = 1
image.show(img1)
return img1
def measure(im, debug=False):
im2 = image.max_size(im, (800, 600))
b,g,r = image.split(im2)
#cv.EqualizeHist(r,r)
##cv.EqualizeHist(g,g)
##cv.EqualizeHist(b,b)
im2 = image.merge(b,g,r)
eyes = 0
#objs = []
#for cascade in eye_cascades:
# print cascade
# cascade = cv.Load(cascade)
# objs = filter_overlap(detect(im, cascade))
# draw_objects(im, objs, color=cv.RGB(0,255,0))
# eyes += len(objs)
#faces = 0
if debug:
im3 = cv.CloneImage(im2)
faces = []
for cascade in face_cascades:#(face_cascades[0],face_cascades[-1]):
cascade = cv.Load(cascade)
detected_faces = detect(im2, cascade)
faces += detected_faces
if debug:
for i,rect in enumerate(faces):
rect.draw(im3, color=cv.RGB(255,16*i,16*i))
if debug:
image.show(im3, "Faces + Repeats")
faces = filter_overlap(faces)
#print (objs[1], objs[6])
#draw_objects(im2, map(tuple, faces.keys()))
for rect, count in faces.iteritems():
rect.draw(im2, color=cv.RGB(255,0,0))
#print (objs[3],objs[13])
#draw_objects(im, filter_overlap((objs[3],objs[13])))
#objs = []
#for cascade in body_cascades:
# print cascade
# cascade = cv.Load(cascade)
# objs += detect(im, cascade)
#draw_objects(im, filter_overlap(objs), color=cv.RGB(0,0,255))
#objs = []
#for cascade in mouth_cascades:
# print cascade
# cascade = cv.Load(cascade)
# objs += detect(im, cascade)
#draw_objects(im, filter_overlap(objs), color=cv.RGB(255,0,255))
score = 0
for face_rect, count in faces.iteritems():
score += count * 0.25 + 0.15
print faces
if debug:
image.show(im2, "Faces")
return (im2, faces), score
def add(request):
data = request.GET['data']
image=qrcode.make(data)
return HttpResponse(image.show())
def main():
parser = argparse.ArgumentParser(description='Decrypts and authenticates binary content hidden in the lowest bits of a bitmap')
parser.add_argument('-k', '--password', type=str, required=True, metavar='<password>', help='The password used for decryption')
parser.add_argument('-m', '--macpassword', type=str, required=True, metavar='<macpassword>', help='The password used for authenticity and integrity')
parser.add_argument('bitmap', type=file, metavar='<bitmap>', help='The bitmap file to extract the binary file from')
args = parser.parse_args()
password = args.password
macpassword = args.macpassword
bitmap = args.bitmap
key = int(hashlib.sha256(password).hexdigest(), 16) >> 128
ciphertext = image.show(bitmap)
iv = binary.pack(slice(ciphertext, 8))
binary_size = binary.pack(slice(ciphertext, 2))
encrypted = slice(ciphertext, binary_size)
decrypted = bytearray(cfb.decrypt(xtea.encrypt, key, iv, encrypted))
mac, decrypted = binary.pack(decrypted[:32]), decrypted[32:]
check = int(hmac.new(macpassword, str(decrypted), hashlib.sha256).hexdigest(), 16)
if mac == check:
for d in decrypted:
sys.stdout.write(chr(d))
else:
raise Exception('Data has been altered')
def measure(im, debug=False):
gray = image.rgb2gray(im)
size = cv.GetSize(im)
total = float(size[0] * size[1])
edges = image.auto_edges(im)
hue, sat, val = tuple(
map(image.equalize_hist, image.split(image.rgb2hsv(im))))
l, u, v = tuple(map(image.equalize_hist, image.split(image.rgb2luv(im))))
values = []
if debug:
image.show(l, "L")
image.show(val, "Value")
sat = image.threshold(val, 255 - 32) #image.And(val, sat)
if debug:
image.show(sat, "Thresh")
#cv.And(val, l, val)
cv.Sub(l, sat, l)
cv.Set(l, 0, image.dilate(edges, iterations=3))
if debug:
image.show(l, "L - Value")
val = l
g = Grid(cv.GetSize(val))
images = g.split_into(val, 16)
arr = image.cv2array(val)
avgmean, avgstd = arr.mean(), arr.std()
for i in images:
a = image.cv2array(i)
mean, std = abs(a.mean() - avgmean), max(a.std(), 0)
values.append((mean + std))
if debug:
print values
print "AVG", avgmean, avgstd
image.show(val, "Result")
return val, (avgmean, avgstd, len([v for v in values if v > avgstd * 2]))
def measure(im, debug=False):
gray = image.rgb2gray(im)
size = cv.GetSize(im)
total = float(size[0] * size[1])
edges = image.auto_edges(im)
_, sat, val = image.split(image.rgb2hsv(im))
edges = image.auto_edges(im)
l, u, v = tuple(map(image.equalize_hist, image.split(image.rgb2luv(im))))
u, v = tuple(map(image.gaussian, (u, v)))
if debug:
image.show(l, "1. L")
image.show(u, "1. U")
image.show(v, "1. V")
la, ua, va, uva = tuple(map(image.cv2array, (l, u, v, image.And(l, u, v))))
test = image.new_from(gray)
test2 = image.new_from(gray)
cv.Xor(u, v, test)
#cv.AbsDiff(u,v, test2)
if debug:
#cv.Threshold(test, test, 32, 255, cv.CV_THRESH_BINARY)
image.show(test, "2. U Xor V")
#image.show(test2, "TEST 2")
#test = image.dilate(test)
cv.Set(test, 0, image.dilate(edges))
#cv.Set(test, 0, image.invert(image.threshold(sat, threshold=8)))
uv_score = cv.CountNonZero(test) / total
if debug:
image.show(
test, "3. U Xor V - dilate(Edges) - invert(threshold(Saturation))")
arr = image.cv2array(test)
avg_mean, avg_std = arr.mean(), arr.std()
score = uv_score, avg_std
return test, score
def measure(im, debug=False):
gray = image.rgb2gray(im)
size = cv.GetSize(im)
total = float(size[0] * size[1])
edges = image.auto_edges(im)
hue, sat, val = tuple(map(image.equalize_hist, image.split(image.rgb2hsv(im)) ))
l,u,v = tuple(map(image.equalize_hist, image.split(image.rgb2luv(im))))
values = []
if debug:
image.show(l, "L")
image.show(val, "Value")
sat = image.threshold(val,255-32)#image.And(val, sat)
if debug:
image.show(sat, "Thresh")
#cv.And(val, l, val)
cv.Sub(l, sat, l)
cv.Set(l, 0, image.dilate(edges, iterations=3))
if debug:
image.show(l, "L - Value")
val = l
g = Grid(cv.GetSize(val))
images = g.split_into(val, 16)
arr = image.cv2array(val)
avgmean, avgstd = arr.mean(), arr.std()
for i in images:
a = image.cv2array(i)
mean, std = abs(a.mean() - avgmean), max(a.std(), 0)
values.append((mean+std))
if debug:
print values
print "AVG", avgmean, avgstd
image.show(val, "Result")
return val, (avgmean, avgstd, len([v for v in values if v > avgstd*2]))
def measure(im, debug=False):
gray = image.rgb2gray(im)
size = cv.GetSize(im)
total = float(size[0] * size[1])
edges = image.auto_edges(im)
_, sat, val = image.split(image.rgb2hsv(im))
edges = image.auto_edges(im)
l,u,v = tuple(map(image.equalize_hist, image.split(image.rgb2luv(im))))
u,v = tuple(map(image.gaussian, (u,v)))
if debug:
image.show(l, "1. L")
image.show(u, "1. U")
image.show(v, "1. V")
la,ua,va,uva = tuple(map(image.cv2array, (l,u,v, image.And(l,u,v))))
test = image.new_from(gray)
test2 = image.new_from(gray)
cv.Xor(u,v,test)
#cv.AbsDiff(u,v, test2)
if debug:
#cv.Threshold(test, test, 32, 255, cv.CV_THRESH_BINARY)
image.show(test, "2. U Xor V")
#image.show(test2, "TEST 2")
#test = image.dilate(test)
cv.Set(test, 0, image.dilate(edges))
#cv.Set(test, 0, image.invert(image.threshold(sat, threshold=8)))
uv_score = cv.CountNonZero(test) / total
if debug:
image.show(test, "3. U Xor V - dilate(Edges) - invert(threshold(Saturation))")
arr = image.cv2array(test)
avg_mean, avg_std = arr.mean(), arr.std()
score = uv_score, avg_std
return test, score
def detect_skin(im, debug=False):
hsv = image.rgb2hsv(im)
if debug:
image.show(hsv, 'hsv')
h,s,v = image.split(hsv)
if cv.CountNonZero(h) == cv.CountNonZero(s) == 0:
white = image.new_from(im)
cv.Set(white, 255)
return white
if debug:
image.show(h, "Hue")
image.show(s,"sat1")
h_rng = 0, 46
s_rng = 48, 178
h = image.threshold(image.gaussian(h, 5), threshold=h_rng[1], type=cv.CV_THRESH_TOZERO_INV)
h = image.threshold(h, threshold=h_rng[0], type=cv.CV_THRESH_TOZERO)
h = image.threshold(h, threshold=1)
s = image.threshold(image.gaussian(s, 5), threshold=s_rng[1], type=cv.CV_THRESH_TOZERO_INV)
s = image.threshold(s, threshold=s_rng[0], type=cv.CV_THRESH_TOZERO)
if debug:
image.show(s,"sat2")
s = image.threshold(s, threshold=1)
v = image.dilate(image.erode(image.And(s, h)))
#im = image.hsv2rgb(image.merge(h,s,v))
if debug:
image.show(v, "Human")
return image.threshold(v, threshold=1)
def measure(im, debug=False):
gray = image.rgb2gray(im)
_, s, v = image.split(image.rgb2hsv(im))
h = GrayscaleHist(bins=64).use_image(v)
s = GrayscaleHist(bins=64).use_image(s)
scores = [score_hist(h)]
if debug:
image.show(v, "1. Value")
image.show(h.to_img(), "2. Value Histogram")
image.show(s.to_img(), "2. Saturation Histogram")
print score_hist(s)
return (None, scores[0]) # h.to_img(),
def measure(im, blur, noise=None, debug=False):
focus_points = blur[0]
#is_noisy = noise[2]
size = cv.GetSize(im)
npixels = size[0] * size[1]
#if focused_regions is None:
# focused_regions = image.new_from(im)
# cv.Set(focused_regions, 0)
# groups = form_groups(focus_points,
# estimated_size=min(max(int(len(npixels) / 1000), 2), 15))
# #groups = form_groups(points, threshold=max(cv.GetSize(im))/16)
# #print 'groups', len(groups)
# draw_groups(groups, focused_regions)
im2 = cv.CloneImage(im)
g = Grid(cv.GetSize(im2))
if debug:
image.show(g.draw(im2), "Image with Grid + ROI")
roi = image.new_from(im, nChannels=1)
cv.Set(roi, 0)
#g.draw_lines(roi, thickness=int(max(min((size[0] + size[1]) * 1/100.0, 255), 1)))
g.draw_regions(roi)
area = cv.Sum(roi)[0]
(_, face_rects), face_score = faces.measure(im)
face_block = image.new_from(im, nChannels=1)
cv.Set(face_block, 0)
for r in face_rects:
r.draw(face_block, color=cv.RGB(255,255,255), thickness=cv.CV_FILLED)
if debug:
face_roi = cv.CloneImage(im)
cv.Set(face_roi, 0, image.invert(roi))
cv.Set(face_roi, 0, image.invert(image.threshold(face_block, threshold=1)))
image.show(face_block, "Faces in Binary")
image.show(g.draw(face_roi), "Face + ROI")
return (im, (
measure_focused_roi(im, roi, area, focus_points, debug),
#measure_color_roi(im, roi, area, focus_points, debug),
measure_contrast(im, debug),
measure_saturation(im, debug),
faces.measure(im, debug)[1],
))
def measure(im, debug=False):
gray = image.rgb2gray(im)
_, s, v = image.split(image.rgb2hsv(im))
h = GrayscaleHist(bins=64).use_image(v)
s = GrayscaleHist(bins=64).use_image(s)
scores = [score_hist(h)]
if debug:
image.show(v, "1. Value")
image.show(h.to_img(), "2. Value Histogram")
image.show(s.to_img(), "2. Saturation Histogram")
print score_hist(s)
return (
None, #h.to_img(),
scores[0],
)
def get_focus_points(im,debug=False):
edges = image.dilate(image.auto_edges(im))
#d = 1
#sobel = image.sobel(im, xorder=d, yorder=d)
sobel = image.laplace(im)
hsv = image.rgb2hsv(im)
focused = image.And(sobel, edges)
if im.nChannels == 3:
hue, saturation, value = image.split(hsv)
saturation = image.dilate(saturation)
focused = image.And(focused, saturation)
focused = image.threshold(image.dilate(focused), threshold=32)
if debug:
image.show(edges, "1. Edges")
image.show(sobel, "2. Sobel")
if im.nChannels == 3:
image.show(saturation, "3. Saturation")
return focused
def get_focus_points(im, debug=False):
edges = image.dilate(image.auto_edges(im))
#d = 1
#sobel = image.sobel(im, xorder=d, yorder=d)
sobel = image.laplace(im)
hsv = image.rgb2hsv(im)
focused = image.And(sobel, edges)
if im.nChannels == 3:
hue, saturation, value = image.split(hsv)
saturation = image.dilate(saturation)
focused = image.And(focused, saturation)
focused = image.threshold(image.dilate(focused), threshold=32)
if debug:
image.show(edges, "1. Edges")
image.show(sobel, "2. Sobel")
if im.nChannels == 3:
image.show(saturation, "3. Saturation")
return focused
def main(progname, *args):
parser = OptionParser()
parser.add_option("-f", "--file", dest="filename", default=None,
help="analyze a given FILE ending in .jpg or .jpeg", metavar="FILE")
parser.add_option("-i", "--imageset", dest="imageset", default=None,
help="Runs on a predefined set of algorithms (li,chow,china,custom)")
parser.add_option("-d", "--debug", dest="debug", action="store_true", default=False,
help="Enable visual debugging.")
parser.add_option("-t", "--type", dest="type", default="all",
help="Specifies the type of feature to debug. Defaults to all.")
(options, args) = parser.parse_args(list(args))
if options.imageset:
if options.imageset == 'li':
process(li_dir)
return 0
elif options.imageset == 'chow':
process(chow_dir)
return 0
elif options.imageset == 'china':
process(china_dir)
return 0
elif options.imageset == 'custom':
process()
return 0
if not options.filename:
print "Please specify a file (--file) or image set (--imageset)."
return 1
if not options.debug:
process([load_image(options.filename)])
return 0
if options.filename.startswith('data/'):
options.filename = options.filename[len('data/'):]
tdata = load_image(options.filename)
kind = options.type.lower()
size = None #(320,240,'crop') # (0.5, 0.5, 'resize-p')
if size is None:
im = tdata.load()
elif size[-1] == 'crop':
im = image.random_cropped_region(tdata.load(), size[:2])
elif size[-1] == 'resize':
im = tdata.load(size[:2])
elif size[-1] == 'resize-p':
im = image.resize(tdata.load(), by_percent=size[:2])
else:
raise TypeError, "Invalid image sizing type."
image.show(im, "Image")
#l,u,v = image.split(image.rgb2luv(im))
##cv.Set(l, 128)
##cv.EqualizeHist(l, l)
##cv.EqualizeHist(u, u)
##image.show(image.luv2rgb(image.merge(l,u,v)), "test")
#s = cv.GetSize(im)
#t = image.absDiff(u,v)
#image.show(t, "test")
#print "Test Score:", cv.CountNonZero(t) / float(s[0] * s[1])
##image.show(image.threshold(image.And(u,v), threshold=1), "LUV")
# noise
if kind in ('all','noise'):
noise_img, score = noise.measure(im, debug=True)
#image.show(noise_img, "Noise Result")
print 'Noise Score:', score, noise.boolean(score)
# contrast
if kind in ('all','contrast'):
contrast_img, score = contrast.measure(im, debug=True)
#image.show(contrast_img, "Contrast Result")
print 'Contrast Score:', score, contrast.boolean(score)
# blur
if kind in ('all','blur','composition'):
focused, score = blur.measure(im, debug=kind in ('all','blur'))
#image.show(focused, "Blur Result")
print 'Blur Score:', score, blur.boolean(score)
# composition
if kind in ('all','composition'):
composition_img, score = composition.measure(im,
(focused,score, blur.boolean(score)), debug=True)
print 'Composition Score:', score, composition.boolean(score)
if kind in ('faces',):
result, score = faces.measure(im,debug=True)
print "Face Score:", score, faces.boolean(faces)
#win = CornerTweaker(im)
#win.show()
#_, sat, _ = image.split(image.rgb2hsv(im))
#arr = image.cv2array(sat)
#print arr.mean(), arr.std()
# faces
#im, score = faces.measure(im, debug=True)
#print score, faces.boolean(score)
# composition
#noise_img, score = noise.measure(im, debug=False)
##n = (noise_img, score, noise.boolean(score))
#hulls, score = blur.measure(im, debug=False)
#b = (hulls, score, blur.boolean(score))
#cimg, score = composition.measure(im, b, debug=True)
#print score, composition.boolean(score)
# BLUR
#from time import time
#start = time()
##im2 = image.threshold(image.laplace(im), threshold=75, type=cv.CV_THRESH_TOZERO)
#hulls, score = blur.measure(im, debug=True)
##blur_img, score = blur.measure(im, debug=True)
#end = time()
#print "Time:", (end - start), "seconds"
#image.show(im, "image")
##image.show(noise_img, "Noise Image")
#print score, blur.boolean(score)
#CONTRAST
#_, score = contrast.measure(im, debug=True)
#image.show(im, "Image")
#print score, contrast.boolean(score)
"""
#BLUR
#im2 = image.threshold(image.laplace(im), threshold=75, type=cv.CV_THRESH_TOZERO)
im3, score = blur.measure(im, debug=True)
image.show(im, "image")
image.show(im3, "Focus Mask")
print score, blur.boolean(score)
#plt.show()
"""
#NOISE
#noise_img, score = noise.measure(im, debug=True)
#image.show(noise_img, "Noise")
#print score, noise.boolean(score)
"""
#hwin = ColorHistograms(im)
#hwin.show()
hwin = HistogramWindow(image.rgb2gray(im))
hwin.show()
print cv.GetSize(im), cv.GetSize(im2)
print 'blur', papers.blurry_histogram(im)
#print papers.blurry_histogram(im2)
wind = DerivativeTweaker(im, title="image derivative")
wind.show()
win = EdgeThresholdTweaker(im, title="image edges")
win.show(50)#edge_threshold(im))
#win2 = EdgeThresholdTweaker(im2, title="image resized edges")
#win2.show(edge_threshold(im2))
"""
cv.WaitKey()
cv.DestroyAllWindows()
return 0
saturation = image_name.saturation()
exposedness = image_name.exposedness()
weight = (contrast**w_c) * (saturation**w_s) * (exposedness**
w_e) + 1e-12
# weight = ndimage.gaussian_filter(weight, sigma=(3, 3), order=0)
self.weights.append(weight)
sums = sums + weight
for index in range(self.num_images):
self.weights[index] = self.weights[index] / sums
return self.weights
def result_exposure(self, w_c=1, w_s=1, w_e=1):
"Return the Exposure Fusion image with Naive method"
self.get_weights_map(w_c, w_s, w_e)
self.result_image = np.zeros(self.shape)
for canal in range(3):
for index in range(self.num_images):
self.result_image[:, :, canal] += self.weights[
index] * self.images[index].array[:, :, canal]
self.result_image[self.result_image < 0] = 0
self.result_image[self.result_image > 1] = 1
return self.result_image
if __name__ == "__main__":
names = [line.rstrip('\n') for line in open('list_jpeg_test.txt')]
W = WeightsMap("mask", names)
im = W.result_exposure(1, 1, 1)
image.show(im)
misc.imsave("res/mask_naive.jpg", im)
dest='w_s',
type=float,
default=1.0,
help='Exponent of the saturation')
parser.add_argument('-we',
dest='w_e',
type=float,
default=1.0,
help='Exponent of the exposedness')
args = parser.parse_args()
params = vars(args) # convert to ordinary dict
names = [line.rstrip('\n') for line in open(params['names'])]
folder = params['folder']
height_pyr = params['height_pyr']
w_c = params['w_c']
w_s = params['w_s']
w_e = params['w_e']
# Naive Fusion
W = naivefusion.WeightsMap(folder, names)
res_naive = W.result_exposure(w_c, w_s, w_e)
image.show(res_naive)
# Laplacian Fusion
lap = laplacianfusion.LaplacianMap(folder, names, n=height_pyr)
res_lap = lap.result_exposure(w_c, w_s, w_e)
image.show(res_lap)
print "laplacian pyramid"
self.get_laplacian_pyramid_images()
result_pyramid = []
for floor in range(self.height_pyr):
print 'floor ', floor
result_floor = np.zeros(self.laplacian_pyramid[0][floor].shape)
for index in range(self.num_images):
print 'image ', index
for canal in range(3):
result_floor[:, :, canal] += self.laplacian_pyramid[index][
floor][:, :,
canal] * self.weights_pyramid[index][floor]
result_pyramid.append(result_floor)
# Get the image from the Laplacian pyramid
self.result_image = result_pyramid[-1]
for floor in range(self.height_pyr - 2, -1, -1):
print 'floor ', floor
self.result_image = result_pyramid[floor] + utils.Expand(
self.result_image, 1)
self.result_image[self.result_image < 0] = 0
self.result_image[self.result_image > 1] = 1
return self.result_image
if __name__ == "__main__":
names = [line.rstrip('\n') for line in open('list_images.txt')]
lap = LaplacianMap('arno', names, n=6)
res = lap.result_exposure(1, 1, 1)
image.show(res)
misc.imsave("res/arno_3.jpg", res)
def measure_color_roi(im, roi, area, focused_regions, debug=False):
im = cv.CloneImage(im)
g = Grid(cv.GetSize(im))
"""
contours = Contours(image.threshold(focused_regions, threshold=1)).approx_poly()
if debug:
test = image.new_from(im)
cv.Set(test, 0)
for c in contours:
i = 1
while c:
cv.FillPoly(test, [[c[x] for x in range(len(c))]], cv.RGB(0,64*i,0))
c = c.v_next()
i += 1
#contours.draw(test, levels=9)
image.show(test, "Test")
"""
#mask = image.And(image.threshold(focused_regions, threshold=1), roi)
#
#canvas = image.new_from(im, nChannels=1)
#cv.Set(canvas, 0)
#if cv.CountNonZero(mask) <= 1:
# return 0, 0
#contours = Contours(image.dilate(mask)).approx_poly()
#for c in contours:
# i = 1
# while c:
# cv.FillPoly(canvas, [[c[x] for x in range(len(c))]], 255)
# c = c.v_next()
# i += 1
#mask = image.Or(mask, canvas)
#if debug:
# image.show(mask, "MASK")
#
#cv.Set(im, 0, image.invert(mask))
cv.Set(im, 0, image.invert(roi))
#area = cv.CountNonZero(image.threshold(im, threshold=1))
if debug:
image.show(g.draw(im,thickness=2), "Image + ROI + Focus point mask")
scores = []
im = image.rgb2gray(im)
#canvas = image.And(plane, roi)
quadrants = g.split_in_four(im)
hist = []
for q,quad in enumerate(quadrants):
#scores.append(cv.Sum(quad)[0] / float(area/4))
h = GrayscaleHist(value_range=(1,255)).use_image(quad)
#image.show(h.to_img(), ['gray', 'red','green','blue'][i] + ' in ' + str(q))
hist.append(h.to_array())
scores = []
excluded_points = set([(2, 1), (3, 0)])
for i,h1 in enumerate(hist):
for j,h2 in enumerate(hist):
if i <= j or (i,j) in excluded_points:
continue
h = abs(h2-h1)
ht = GrayscaleHist(value_range=(0,255)).use_array_as_hist(h)
scores.append((h[5:].mean(), h[5:].std()))
means = max([x[0] for x in scores])
stddevs = max([x[1] for x in scores])
return means/255.0, stddevs/255.0
def main(progname, *args):
parser = OptionParser()
parser.add_option("-f",
"--file",
dest="filename",
default=None,
help="analyze a given FILE ending in .jpg or .jpeg",
metavar="FILE")
parser.add_option(
"-i",
"--imageset",
dest="imageset",
default=None,
help="Runs on a predefined set of algorithms (li,chow,china,custom)")
parser.add_option("-d",
"--debug",
dest="debug",
action="store_true",
default=False,
help="Enable visual debugging.")
parser.add_option(
"-t",
"--type",
dest="type",
default="all",
help="Specifies the type of feature to debug. Defaults to all.")
(options, args) = parser.parse_args(list(args))
if options.imageset:
if options.imageset == 'li':
process(li_dir)
return 0
elif options.imageset == 'chow':
process(chow_dir)
return 0
elif options.imageset == 'china':
process(china_dir)
return 0
elif options.imageset == 'custom':
process()
return 0
if not options.filename:
print "Please specify a file (--file) or image set (--imageset)."
return 1
if not options.debug:
process([load_image(options.filename)])
return 0
if options.filename.startswith('data/'):
options.filename = options.filename[len('data/'):]
tdata = load_image(options.filename)
kind = options.type.lower()
size = None #(320,240,'crop') # (0.5, 0.5, 'resize-p')
if size is None:
im = tdata.load()
elif size[-1] == 'crop':
im = image.random_cropped_region(tdata.load(), size[:2])
elif size[-1] == 'resize':
im = tdata.load(size[:2])
elif size[-1] == 'resize-p':
im = image.resize(tdata.load(), by_percent=size[:2])
else:
raise TypeError, "Invalid image sizing type."
image.show(im, "Image")
#l,u,v = image.split(image.rgb2luv(im))
##cv.Set(l, 128)
##cv.EqualizeHist(l, l)
##cv.EqualizeHist(u, u)
##image.show(image.luv2rgb(image.merge(l,u,v)), "test")
#s = cv.GetSize(im)
#t = image.absDiff(u,v)
#image.show(t, "test")
#print "Test Score:", cv.CountNonZero(t) / float(s[0] * s[1])
##image.show(image.threshold(image.And(u,v), threshold=1), "LUV")
# noise
if kind in ('all', 'noise'):
noise_img, score = noise.measure(im, debug=True)
#image.show(noise_img, "Noise Result")
print 'Noise Score:', score, noise.boolean(score)
# contrast
if kind in ('all', 'contrast'):
contrast_img, score = contrast.measure(im, debug=True)
#image.show(contrast_img, "Contrast Result")
print 'Contrast Score:', score, contrast.boolean(score)
# blur
if kind in ('all', 'blur', 'composition'):
focused, score = blur.measure(im, debug=kind in ('all', 'blur'))
#image.show(focused, "Blur Result")
print 'Blur Score:', score, blur.boolean(score)
# composition
if kind in ('all', 'composition'):
composition_img, score = composition.measure(
im, (focused, score, blur.boolean(score)), debug=True)
print 'Composition Score:', score, composition.boolean(score)
if kind in ('faces', ):
result, score = faces.measure(im, debug=True)
print "Face Score:", score, faces.boolean(faces)
#win = CornerTweaker(im)
#win.show()
#_, sat, _ = image.split(image.rgb2hsv(im))
#arr = image.cv2array(sat)
#print arr.mean(), arr.std()
# faces
#im, score = faces.measure(im, debug=True)
#print score, faces.boolean(score)
# composition
#noise_img, score = noise.measure(im, debug=False)
##n = (noise_img, score, noise.boolean(score))
#hulls, score = blur.measure(im, debug=False)
#b = (hulls, score, blur.boolean(score))
#cimg, score = composition.measure(im, b, debug=True)
#print score, composition.boolean(score)
# BLUR
#from time import time
#start = time()
##im2 = image.threshold(image.laplace(im), threshold=75, type=cv.CV_THRESH_TOZERO)
#hulls, score = blur.measure(im, debug=True)
##blur_img, score = blur.measure(im, debug=True)
#end = time()
#print "Time:", (end - start), "seconds"
#image.show(im, "image")
##image.show(noise_img, "Noise Image")
#print score, blur.boolean(score)
#CONTRAST
#_, score = contrast.measure(im, debug=True)
#image.show(im, "Image")
#print score, contrast.boolean(score)
"""
#BLUR
#im2 = image.threshold(image.laplace(im), threshold=75, type=cv.CV_THRESH_TOZERO)
im3, score = blur.measure(im, debug=True)
image.show(im, "image")
image.show(im3, "Focus Mask")
print score, blur.boolean(score)
#plt.show()
"""
#NOISE
#noise_img, score = noise.measure(im, debug=True)
#image.show(noise_img, "Noise")
#print score, noise.boolean(score)
"""
#hwin = ColorHistograms(im)
#hwin.show()
hwin = HistogramWindow(image.rgb2gray(im))
hwin.show()
print cv.GetSize(im), cv.GetSize(im2)
print 'blur', papers.blurry_histogram(im)
#print papers.blurry_histogram(im2)
wind = DerivativeTweaker(im, title="image derivative")
wind.show()
win = EdgeThresholdTweaker(im, title="image edges")
win.show(50)#edge_threshold(im))
#win2 = EdgeThresholdTweaker(im2, title="image resized edges")
#win2.show(edge_threshold(im2))
"""
cv.WaitKey()
cv.DestroyAllWindows()
return 0
def measure(im, debug=False):
gray = image.rgb2gray(im)
size = cv.GetSize(im)
total = float(size[0] * size[1])
l = image.sub(gray, image.gaussian(gray, 5))
l2 = image.sub(gray, image.gaussian(gray, 9))
edges = image.dilate(image.auto_edges(im, percentage=0.2))
if debug:
image.show(image.threshold(l, threshold=1), "Before Edge Removal (kernel=5)")
image.show(image.threshold(l2, threshold=1), "Before Edge Removal (kernel=9)")
cv.Set(l, 0, image.threshold(edges, threshold=1))
cv.Set(l2, 0, image.threshold(edges, threshold=1))
l = image.threshold(l, threshold=1)
l2 = image.threshold(l2, threshold=1)
if debug:
image.show(image.threshold(edges, threshold=1), "Edges")
image.show(l, "After Edge Removal (kernel=5)")
image.show(l2, "After Edge Removal (kernel=9)")
noise2 = image.new_from(gray)
cv.EqualizeHist(gray, noise2)
cv.AbsDiff(noise2, gray, noise2)
cv.Set(noise2, 0, image.threshold(image.sobel(im, xorder=2, yorder=2), threshold=4))
diff = image.cv2array(noise2)
if debug:
image.show(noise2, "DIFF")
print "M", diff.mean(), "S", diff.std()
diff_stat = (diff.mean(), diff.std())
percent_noise = cv.CountNonZero(noise2) / total
if debug:
image.show(noise2, "NOISE2")
# magical, I don't understand how this works
_, sat, _ = image.split(image.rgb2hsv(im))
edges = image.auto_edges(im)
l,u,v = tuple(map(image.equalize_hist, image.split(image.rgb2luv(im))))
u,v = tuple(map(image.gaussian, (u,v)))
if debug:
image.show(l, "1. L")
image.show(u, "1. U")
image.show(v, "1. V")
la,ua,va,uva = tuple(map(image.cv2array, (l,u,v, image.And(l,u,v))))
test = image.new_from(gray)
test2 = image.new_from(gray)
cv.Xor(u,v,test)
if debug:
image.show(test, "2. U Xor V")
cv.Set(test, 0, image.dilate(edges))
#cv.Set(test, 0, image.invert(image.threshold(sat, threshold=8)))
uv_score = cv.CountNonZero(test) / total
if debug:
image.show(test, "3. U Xor V - dilate(Edges) - invert(threshold(Saturation))")
g = Grid(size)
images = map(image.cv2array, g.split_into(test, 6))
arr = image.cv2array(test)
avg_mean, avg_std = arr.mean(), arr.std()
#ms = [(a.mean(), a.std()) for a in images]
#min_mean = min_std = 255
#max_mean = max_std = 0
#for m,s in ms:
# min_mean = min(min_mean, m)
# min_std = min(min_std, s)
# max_mean = max(max_mean, m)
# max_std = max(max_std, s)
#if debug:
# print min_mean, min_std
# print avg_mean, avg_std
# print max_mean, max_std
#
#score = uv_score, min_mean, avg_mean, avg_std, max_mean
uv_score = uv_score, avg_std
score = cv.CountNonZero(l) / total, cv.CountNonZero(l2) / total, \
diff_stat[0], diff_stat[1], uv_score
return l, score
def measure(im, debug=False):
size = cv.GetSize(im)
npixels = size[0] * size[1]
#print 'np', npixels
focused = get_focus_points(im, debug)
points = convert_to_points(focused)
if debug:
print "\t" + str(
len(points)), '/', npixels, '=', len(points) / float(npixels)
print "\tlen(points) =", len(points)
image.show(focused, "4. Focused Points")
saturation_score = 0
if not image.is_grayscale(im):
edges = image.auto_edges(im)
_, saturation, _ = image.split(image.rgb2hsv(im))
if debug:
image.show(saturation, "5. Saturation")
#saturation = image.laplace(image.gaussian(saturation, 3))
saturation = image.invert(saturation)
mask = image.invert(image.threshold(im, threshold=16))
if debug:
image.show(saturation, "5.3. Laplace of Saturation")
cv.Set(saturation, 0, mask)
cv.Set(saturation, 0, focused)
if debug:
image.show(mask,
"5.6. Mask(focused AND invert(threshold(im, 16)))")
image.show(saturation, "6. Set(<5.3>, 0, <5.6>)")
saturation_score = cv.Sum(saturation)[0] / float(npixels * 255)
print "\tSaturation Score:", saturation_score
# light exposure
h, s, v = image.split(image.rgb2hsv(im))
if debug:
image.show(h, "7. Hue")
image.show(s, "7. Saturation")
image.show(v, "7. Value")
diff = cv.CloneImage(v)
cv.Set(diff, 0, image.threshold(s, threshold=16))
diff = image.dilate(diff, iterations=10)
if debug:
thres_s = image.threshold(s, threshold=16)
image.show(thres_s, "8.3. Mask(threshold(<7.Saturation>, 16))")
image.show(diff, "8.6. Dilate(Set(<7.Value>, 0, <8.3>), 10)")
cdiff = cv.CountNonZero(diff)
if cdiff > 0 and cdiff / float(npixels) > 0.01:
test = cv.CloneImage(v)
cv.Set(test, 0, image.invert(diff))
s = cv.Sum(test)[0] / float(cdiff * 255)
if debug:
print '\tLight Exposure Score:', s
else:
s = 0
if image.is_grayscale(im):
return focused, (1, 1, 1, saturation_score, s)
# we want to short circuit ASAP to avoid doing KMeans 50% of the image's pixels
if len(points) > npixels / 2:
return focused, (1, 1, 1, saturation_score, s)
# we're so blurry we don't have any points!
if len(points) < 1:
return focused, (0, 0, 0, saturation_score, s)
if debug:
im2 = cv.CloneImage(im)
focused_regions = image.new_from(im)
cv.Set(focused_regions, 0)
r = lambda x: random.randrange(1, x)
groups = form_groups(points,
estimated_size=min(max(int(len(points) / 1000), 2),
15))
#groups = form_groups(points, threshold=max(cv.GetSize(im))/16)
#print 'groups', len(groups)
hulls = draw_groups(groups, focused_regions)
focused_regions = image.threshold(focused_regions,
threshold=32,
type=cv.CV_THRESH_TOZERO)
min_area = npixels * 0.0005
densities = [h.points_per_pixel() for h in hulls if h.area() >= min_area]
if debug:
#image.show(focused, "Focused Points")
image.show(focused_regions, "9. Focused Regions from <4>")
cv.Sub(
im2,
image.gray2rgb(
image.invert(image.threshold(focused_regions, threshold=1))),
im2)
image.show(im2, "10. threshold(<9>)")
focused_regions = image.rgb2gray(focused_regions)
densities = array(densities)
c = cv.CountNonZero(focused_regions)
c /= float(npixels)
score = (c, densities.mean(), densities.std(), saturation_score, s)
return focused, score
def gcs_matching(L, R, Seeds=None, tau=TAU, mu=MU, show=True):
# NEW: Add padding...
if (ROLL == False):
L = zero_pad(L, DISPARITY)
R = zero_pad(R, DISPARITY)
# ...
(height, width) = shape = L.shape
# ...
auxilary = numpy.nan * numpy.ones([height, width, DISPARITY_SPAN])
Tau = Tau_initialize(shape)
Tau_count = 0
# Best costs
CC = 0 - numpy.inf * numpy.ones(shape)
CC_prime = 0 - numpy.inf * numpy.ones(shape)
# Best disparities
DD = 0 - (1 + DISPARITY) * numpy.ones(shape)
DD_prime = 0 - (1 + DISPARITY) * numpy.ones(shape)
# Handle preemptive matching
if Seeds is None:
Seeds = preemptive_match(auxilary, L, R)
# Setup visuals
if show:
import matplotlib.pyplot as pyplot
figure, axis = pyplot.subplots()
pyplot.plot()
pyplot.hold(True)
print("GCS Algorithm...")
while True:
sys.stdout.write('\r')
sys.stdout.write("[S Size]: %d" % Seeds.qsize())
sys.stdout.write('\t')
sys.stdout.write("[Tau Size]: %d" % Tau_count)
sys.stdout.flush()
if Seeds.empty(): break
(_, ss) = Seeds.get()
for qq in get_best_neighbours(auxilary, L, R, ss):
(vv, uu, uu_prime) = qq
cc = get_similarity(auxilary, L, R, qq)
if (cc < tau): continue
if Tau_contains(Tau, qq): continue
if (cc + mu < numpy.min([CC[vv][uu], CC_prime[vv][uu_prime]])):
continue
Tau_insert(Tau, qq)
Tau_count += 1
Seeds.put((0 - cc, qq))
if (CC[vv][uu] < cc):
CC[vv][uu] = cc
DD[vv][uu] = get_disparity(qq)
if (CC_prime[vv][uu_prime] < cc):
CC_prime[vv][uu_prime] = cc
DD_prime[vv][uu_prime] = get_disparity(qq)
if show and (Tau_count % 1000) == 0:
image.show(DD, block=False)
figure.canvas.draw()
axis.cla()
sys.stdout.write('\n')
return (DD, DD_prime, CC, CC_prime)
#numpy.random.shuffle(dd_range) # Unnecessary
cc = [ get_similarity(L_image, R_image, (yy_range[ii], xx_range[ii], xx_range[ii]+dd), auxilary) for dd in dd_range ]
kk = numpy.argmax(cc)
Seeds.put( (0-cc[kk], (yy_range[ii], xx_range[ii], xx_range[ii]+dd_range[kk]) ) )
print(Seeds.qsize(), Seeds_count)
# ...
(DD, DD_prime, CC, CC_prime) = gcs_matching(L_image, R_image, Seeds, auxilary)
pickle.dump(DD, open("DD", "wb"))
pickle.dump(DD_prime, open("DD_prime", "wb"))
pickle.dump(CC, open("CC", "wb"))
pickle.dump(CC_prime, open("CC_prime", "wb"))
exit()
else:
DD = pickle.load(open("DD", "rb"))
DD_prime = pickle.load(open("DD_prime", "rb"))
CC = pickle.load(open("CC", "rb"))
CC_prime = pickle.load(open("CC_prime", "rb"))
(height, width) = shape = DD.shape
image.show(DD)
image.show(DD_prime)
#
"""
RGB color image
"""
from scipy import misc
import numpy as np
import image as img
class rgb_image(img.image):
"""
"""
def __init__(self, w=0, h=0, fileName=None):
"""
Construction d'une image de width x height pixels contenant trois channels R,G et B
"""
if None == fileName:
img.image.__init__(self, w, h, 3)
else:
data = misc.imread(fileName)
img.image.__init__(self, data.shape[0], data.shape[1], 3)
self.__pixels__ = data[:, :, 0:3] / 255.
if __name__ == '__main__':
img = rgb_image(fileName='lena_rgb.png')
img.pixels = np.ones((img.height, img.width, 3), np.double) - img.pixels
img.show()
img.save("inverse_rgb_lena.png")
l = cv.CreateImage((256,256), cv.IPL_DEPTH_8U, 1)
cv.Set(l, 255)
u = image.new_from(l)
v = image.new_from(l)
cv.Set(u, 0)
cv.Set(v, 0)
size = cv.GetSize(l)
print size
for x in range(256):
for y in range(size[1]):
cv.Set2D(u, y, x, x)
cv.Set2D(v, 255-x, min(y, 255), x)
image.show(u, "U")
image.show(v, "V")
rgb = image.luv2rgb(image.merge(l,u,v))
r,g,b = image.split(rgb)
#xor = image.threshold(image.Xor(u,v), 0, cv.CV_THRESH_BINARY)
xor = image.Xor(u,v)
cv.Threshold(xor, xor, 16, 255, cv.CV_THRESH_TOZERO)
image.show(rgb, "RGB")
image.show(xor, "Xor")
#cv.Sub(rgb, image.gray2rgb(image.invert(xor)), rgb)
_, sat, _ = image.split(image.rgb2hsv(rgb))
image.show(sat, 'Saturation')
#cv.Set(xor, 0, image.invert(image.threshold(sat, threshold=4)))
aspect = float(MAX_WIDTH)/width
L = arrays.resample(L, [ int(aspect*x) for x in L.shape ])
R = arrays.resample(R, [ int(aspect*x) for x in R.shape ])
print("Left shape: " + str(L.shape))
print("Right shape: " + str(R.shape))
print("Compute disparity map with LEFT image as reference")
if (os.path.exists(L_disparity_map_path)):
L_disparity_map = pickle.load(open(L_disparity_map_path, "rb"))
else:
L_disparity_map = block_matching(L, R)
pickle.dump(L_disparity_map, open(L_disparity_map_path, "wb"))
image.show(L_disparity_map, title = "Left Disparity Map")
print("Compute disparity map with RIGHT image as reference")
if (os.path.exists(R_disparity_map_path)):
R_disparity_map = pickle.load(open(R_disparity_map_path, "rb"))
else:
R_disparity_map = block_matching(R, L)
pickle.dump(R_disparity_map, open(R_disparity_map_path, "wb"))
image.show(R_disparity_map, title = "Right Disparity Map")
print("Remove any inconsistency between both images")
depth_map = remove_inconsistency(L_disparity_map, R_disparity_map, empty_value = 8)
depth_map = 0-depth_map
l = cv.CreateImage((256, 256), cv.IPL_DEPTH_8U, 1)
cv.Set(l, 255)
u = image.new_from(l)
v = image.new_from(l)
cv.Set(u, 0)
cv.Set(v, 0)
size = cv.GetSize(l)
print size
for x in range(256):
for y in range(size[1]):
cv.Set2D(u, y, x, x)
cv.Set2D(v, 255 - x, min(y, 255), x)
image.show(u, "U")
image.show(v, "V")
rgb = image.luv2rgb(image.merge(l, u, v))
r, g, b = image.split(rgb)
#xor = image.threshold(image.Xor(u,v), 0, cv.CV_THRESH_BINARY)
xor = image.Xor(u, v)
cv.Threshold(xor, xor, 16, 255, cv.CV_THRESH_TOZERO)
image.show(rgb, "RGB")
image.show(xor, "Xor")
#cv.Sub(rgb, image.gray2rgb(image.invert(xor)), rgb)
_, sat, _ = image.split(image.rgb2hsv(rgb))
image.show(sat, 'Saturation')
#cv.Set(xor, 0, image.invert(image.threshold(sat, threshold=4)))
def block_matching(
L_image,
R_image,
disparity = DISPARITY,
window = WINDOW,
roll = True,
cost_default = numpy.inf,
cost_function = costs.ssd,
censor = True,
censor_threshold = 2.0,
show = True):
# Brute-force matching algorithm!
L_image = arrays.as_bytes(L_image)
R_image = arrays.as_bytes(R_image)
# ...
shape = (height, width) = L_image.shape
window_span = 1 + 2*window
disparity_span = 1 + 2*disparity
disparity_range = (0-disparity) + numpy.array(range(disparity_span))
disparity_map = (0-disparity) * numpy.ones(shape)
if roll:
yy_range = xrange(height)
xx_range = xrange(width)
else:
yy_range = xrange(window, height-window)
xx_range = xrange(window+disparity, width-window-disparity)
if show:
figure, axis = pyplot.subplots()
pyplot.plot()
pyplot.hold(True)
for yy in yy_range:
percent_complete = numpy.floor(100*(float(1+yy)/height))
sys.stdout.write("\rProgress: %d%%" % percent_complete)
sys.stdout.flush()
if show:
image.show(disparity_map, block = False)
figure.canvas.draw()
axis.cla()
for xx in xx_range:
L_window = image.neighbours(L_image, yy, xx, size = window, roll = roll)
if (censor): # Censorship: Remove areas of low texture...
# Censor variance along the horizontal scanline, try: { 2, 8, 16, 32, 48 }
# IMPORTANT:
# Rows (xx) is axis=1, ie. numpy.sum(W, axis=1) <=> W[:][0]+W[:][1]+W[:][2]
# Columns (yy) is axis=0, ie. numpy.sum(W, axis=0) <=> W[0][:]+W[1][:]+W[2][:]
scanline_mean = numpy.mean(L_window, axis = 0) # Horizontal scanline mean
scanline_variance = numpy.mean((L_window - scanline_mean)**2) # Horizontal scanline mean
if (scanline_variance < censor_threshold):
continue # Ensure exture quality
# Find the best disparity match...
''' #This one-liner doesn't seem to speed things up...
disparity_map[yy][xx] = (0-disparity) + numpy.array([
cost_function(L_window, image.neighbours(R, yy, xx+dd, size = window, roll = roll))
for dd in disparity_range
]).argmin()
'''
cc_best = cost_default
for dd in disparity_range:
R_window = image.neighbours(R_image, yy, xx+dd, size = window, roll = roll)
cc = cost_function(L_window, R_window)
if (cc < cc_best):
cc_best = cc
disparity_map[yy][xx] = dd
sys.stdout.write("\r")
return disparity_map
def measure(im, debug=False):
size = cv.GetSize(im)
npixels = size[0] * size[1]
#print 'np', npixels
focused = get_focus_points(im, debug)
points = convert_to_points(focused)
if debug:
print "\t"+str(len(points)), '/', npixels, '=', len(points) / float(npixels)
print "\tlen(points) =", len(points)
image.show(focused, "4. Focused Points")
saturation_score = 0
if not image.is_grayscale(im):
edges = image.auto_edges(im)
_, saturation, _ = image.split(image.rgb2hsv(im))
if debug:
image.show(saturation, "5. Saturation")
#saturation = image.laplace(image.gaussian(saturation, 3))
saturation = image.invert(saturation)
mask = image.invert(image.threshold(im, threshold=16))
if debug:
image.show(saturation, "5.3. Laplace of Saturation")
cv.Set(saturation, 0, mask)
cv.Set(saturation, 0, focused)
if debug:
image.show(mask, "5.6. Mask(focused AND invert(threshold(im, 16)))")
image.show(saturation, "6. Set(<5.3>, 0, <5.6>)")
saturation_score = cv.Sum(saturation)[0] / float(npixels * 255)
print "\tSaturation Score:", saturation_score
# light exposure
h,s,v = image.split(image.rgb2hsv(im))
if debug:
image.show(h, "7. Hue")
image.show(s, "7. Saturation")
image.show(v, "7. Value")
diff = cv.CloneImage(v)
cv.Set(diff, 0, image.threshold(s, threshold=16))
diff = image.dilate(diff, iterations=10)
if debug:
thres_s = image.threshold(s, threshold=16)
image.show(thres_s, "8.3. Mask(threshold(<7.Saturation>, 16))")
image.show(diff, "8.6. Dilate(Set(<7.Value>, 0, <8.3>), 10)")
cdiff = cv.CountNonZero(diff)
if cdiff > 0 and cdiff / float(npixels) > 0.01:
test = cv.CloneImage(v)
cv.Set(test, 0, image.invert(diff))
s = cv.Sum(test)[0] / float(cdiff * 255)
if debug:
print '\tLight Exposure Score:', s
else:
s = 0
if image.is_grayscale(im):
return focused, (1, 1, 1, saturation_score, s)
# we want to short circuit ASAP to avoid doing KMeans 50% of the image's pixels
if len(points) > npixels/2:
return focused, (1, 1, 1, saturation_score, s)
# we're so blurry we don't have any points!
if len(points) < 1:
return focused, (0, 0, 0, saturation_score, s)
if debug:
im2 = cv.CloneImage(im)
focused_regions = image.new_from(im)
cv.Set(focused_regions, 0)
r = lambda x: random.randrange(1, x)
groups = form_groups(points,
estimated_size=min(max(int(len(points) / 1000), 2), 15))
#groups = form_groups(points, threshold=max(cv.GetSize(im))/16)
#print 'groups', len(groups)
hulls = draw_groups(groups, focused_regions)
focused_regions = image.threshold(focused_regions, threshold=32, type=cv.CV_THRESH_TOZERO)
min_area = npixels * 0.0005
densities = [h.points_per_pixel() for h in hulls if h.area() >= min_area]
if debug:
#image.show(focused, "Focused Points")
image.show(focused_regions, "9. Focused Regions from <4>")
cv.Sub(im2, image.gray2rgb(image.invert(image.threshold(focused_regions, threshold=1))), im2)
image.show(im2, "10. threshold(<9>)")
focused_regions = image.rgb2gray(focused_regions)
densities = array(densities)
c = cv.CountNonZero(focused_regions)
c /= float(npixels)
score = (c, densities.mean(), densities.std(), saturation_score, s)
return focused, score
aspect = float(MAX_WIDTH) / width
L = arrays.resample(L, [int(aspect * x) for x in L.shape])
R = arrays.resample(R, [int(aspect * x) for x in R.shape])
print("Left shape: " + str(L.shape))
print("Right shape: " + str(R.shape))
print("Compute disparity map with LEFT image as reference")
if (os.path.exists(L_disparity_map_path)):
L_disparity_map = pickle.load(open(L_disparity_map_path, "rb"))
else:
L_disparity_map = block_matching(L, R)
pickle.dump(L_disparity_map, open(L_disparity_map_path, "wb"))
image.show(L_disparity_map, title="Left Disparity Map")
print("Compute disparity map with RIGHT image as reference")
if (os.path.exists(R_disparity_map_path)):
R_disparity_map = pickle.load(open(R_disparity_map_path, "rb"))
else:
R_disparity_map = block_matching(R, L)
pickle.dump(R_disparity_map, open(R_disparity_map_path, "wb"))
image.show(R_disparity_map, title="Right Disparity Map")
print("Remove any inconsistency between both images")
depth_map = remove_inconsistency(L_disparity_map,
R_disparity_map,
from flask import Flask, render_template, request
import requests
from bs4 import BeautifulSoup
import datetime
import selenium
from selenium import webdriver
import image
#https://www.weather.go.kr/mini/marine/wavemodel_c.jsp?prefix=kim_cww3_%5BAREA%5D_wdpr_&area=jeju&tm=2020.04.28.09&ftm=s000&newTm=2020.04.28.09&x=4&y=10
a = input("날짜")
url = "https://www.weather.go.kr/mini/marine/wavemodel_c.jsp?prefix=kim_cww3_%5BAREA%5D_wdpr_&area=jeju&tm=" + a
driver = webdriver.Chrome()
driver.implicitly_wait(3)
driver.get(url)
soup = BeautifulSoup(driver.page_source, "html.parser")
for i in soup.select("#chart_image"):
src = i.find("img")['src']
image = image.open(src)
image.show()
def block_matching(L_image,
R_image,
disparity=DISPARITY,
window=WINDOW,
roll=True,
cost_default=numpy.inf,
cost_function=costs.ssd,
censor=True,
censor_threshold=2.0,
show=True):
# Brute-force matching algorithm!
L_image = arrays.as_bytes(L_image)
R_image = arrays.as_bytes(R_image)
# ...
shape = (height, width) = L_image.shape
window_span = 1 + 2 * window
disparity_span = 1 + 2 * disparity
disparity_range = (0 - disparity) + numpy.array(range(disparity_span))
disparity_map = (0 - disparity) * numpy.ones(shape)
if roll:
yy_range = xrange(height)
xx_range = xrange(width)
else:
yy_range = xrange(window, height - window)
xx_range = xrange(window + disparity, width - window - disparity)
if show:
figure, axis = pyplot.subplots()
pyplot.plot()
pyplot.hold(True)
for yy in yy_range:
percent_complete = numpy.floor(100 * (float(1 + yy) / height))
sys.stdout.write("\rProgress: %d%%" % percent_complete)
sys.stdout.flush()
if show:
image.show(disparity_map, block=False)
figure.canvas.draw()
axis.cla()
for xx in xx_range:
L_window = image.neighbours(L_image,
yy,
xx,
size=window,
roll=roll)
if (censor): # Censorship: Remove areas of low texture...
# Censor variance along the horizontal scanline, try: { 2, 8, 16, 32, 48 }
# IMPORTANT:
# Rows (xx) is axis=1, ie. numpy.sum(W, axis=1) <=> W[:][0]+W[:][1]+W[:][2]
# Columns (yy) is axis=0, ie. numpy.sum(W, axis=0) <=> W[0][:]+W[1][:]+W[2][:]
scanline_mean = numpy.mean(L_window,
axis=0) # Horizontal scanline mean
scanline_variance = numpy.mean(
(L_window - scanline_mean)**2) # Horizontal scanline mean
if (scanline_variance < censor_threshold):
continue # Ensure exture quality
# Find the best disparity match...
''' #This one-liner doesn't seem to speed things up...
disparity_map[yy][xx] = (0-disparity) + numpy.array([
cost_function(L_window, image.neighbours(R, yy, xx+dd, size = window, roll = roll))
for dd in disparity_range
]).argmin()
'''
cc_best = cost_default
for dd in disparity_range:
R_window = image.neighbours(R_image,
yy,
xx + dd,
size=window,
roll=roll)
cc = cost_function(L_window, R_window)
if (cc < cc_best):
cc_best = cc
disparity_map[yy][xx] = dd
sys.stdout.write("\r")
return disparity_map
_transformation = seismic.parse_transformation(_seismic, 15 * 1000,
int(2.5 * 1000), 10) # 2.5
#test shit begin
test_data, test_match = test_trapeziums()
_geo = Geo(Well(test_data[0], test_data[2]),
Well(test_data[1], test_data[3]), _transformation.width,
_transformation.left_well_intent,
_transformation.right_well_intent)
_match = Match(
[match[1][0] for match in reversed(test_match)],
[match[0][0] + 1 for match in reversed(test_match)],
[match[1][1] for match in reversed(test_match)],
[match[0][1] + 1 for match in reversed(test_match)],
)
#test shit end
_image = image.create(_geo.height, _geo.width)
# paint.lines(_image, _geo, _match)
paint.fill(_image, _geo, _match)
paint.wells(_image, _geo)
_base_image = image.create(_geo.height, _geo.width)
paint.wells(_base_image, _geo)
paint.lines(_base_image, _geo, _match)
_result = image.create(_transformation.height, _transformation.width)
seismic.paint_transformation(_result, _transformation, _image)
image.show(_image, _base_image, _result, paint.bined(_result))
def measure(im, debug=False):
gray = image.rgb2gray(im)
size = cv.GetSize(im)
total = float(size[0] * size[1])
l = image.sub(gray, image.gaussian(gray, 5))
l2 = image.sub(gray, image.gaussian(gray, 9))
edges = image.dilate(image.auto_edges(im, percentage=0.2))
if debug:
image.show(image.threshold(l, threshold=1),
"Before Edge Removal (kernel=5)")
image.show(image.threshold(l2, threshold=1),
"Before Edge Removal (kernel=9)")
cv.Set(l, 0, image.threshold(edges, threshold=1))
cv.Set(l2, 0, image.threshold(edges, threshold=1))
l = image.threshold(l, threshold=1)
l2 = image.threshold(l2, threshold=1)
if debug:
image.show(image.threshold(edges, threshold=1), "Edges")
image.show(l, "After Edge Removal (kernel=5)")
image.show(l2, "After Edge Removal (kernel=9)")
noise2 = image.new_from(gray)
cv.EqualizeHist(gray, noise2)
cv.AbsDiff(noise2, gray, noise2)
cv.Set(noise2, 0,
image.threshold(image.sobel(im, xorder=2, yorder=2), threshold=4))
diff = image.cv2array(noise2)
if debug:
image.show(noise2, "DIFF")
print "M", diff.mean(), "S", diff.std()
diff_stat = (diff.mean(), diff.std())
percent_noise = cv.CountNonZero(noise2) / total
if debug:
image.show(noise2, "NOISE2")
# magical, I don't understand how this works
_, sat, _ = image.split(image.rgb2hsv(im))
edges = image.auto_edges(im)
l, u, v = tuple(map(image.equalize_hist, image.split(image.rgb2luv(im))))
u, v = tuple(map(image.gaussian, (u, v)))
if debug:
image.show(l, "1. L")
image.show(u, "1. U")
image.show(v, "1. V")
la, ua, va, uva = tuple(map(image.cv2array, (l, u, v, image.And(l, u, v))))
test = image.new_from(gray)
test2 = image.new_from(gray)
cv.Xor(u, v, test)
if debug:
image.show(test, "2. U Xor V")
cv.Set(test, 0, image.dilate(edges))
#cv.Set(test, 0, image.invert(image.threshold(sat, threshold=8)))
uv_score = cv.CountNonZero(test) / total
if debug:
image.show(
test, "3. U Xor V - dilate(Edges) - invert(threshold(Saturation))")
g = Grid(size)
images = map(image.cv2array, g.split_into(test, 6))
arr = image.cv2array(test)
avg_mean, avg_std = arr.mean(), arr.std()
#ms = [(a.mean(), a.std()) for a in images]
#min_mean = min_std = 255
#max_mean = max_std = 0
#for m,s in ms:
# min_mean = min(min_mean, m)
# min_std = min(min_std, s)
# max_mean = max(max_mean, m)
# max_std = max(max_std, s)
#if debug:
# print min_mean, min_std
# print avg_mean, avg_std
# print max_mean, max_std
#
#score = uv_score, min_mean, avg_mean, avg_std, max_mean
uv_score = uv_score, avg_std
score = cv.CountNonZero(l) / total, cv.CountNonZero(l2) / total, \
diff_stat[0], diff_stat[1], uv_score
return l, score
import seismic
from data import *
if __name__ == '__main__':
_seismic = image.load("SeismicScaled.jpg")
_transformation = seismic.parse_transformation(_seismic, 15 * 1000, int(2.5 * 1000), 10)
_geo = Geo(Well.generate_left(), Well.generate_right(), _transformation.width, _transformation.left_well_intent, _transformation.right_well_intent)
_match = Match.generate()
_image = image.create(_geo.height, _geo.width)
# image.show(_seismic)
paint.wells(_image, _geo)
paint.lines(_image, _geo, _match)
paint.fill(_image, _geo, _match)
# _image1 = _image.copy()
# paint.fill(_image1, _geo, _match)
# paint.wells(_image1, _geo)
#
# _, _image2 = cv2.threshold(_image1, 127, 255, cv2.THRESH_BINARY)
#
# print("Showing ...")
# image.show(_image, _image1, _image2)
_result = image.create(_transformation.height, _transformation.width)
seismic.paint_transformation(_result, _transformation, _image)
image.show(_image, paint.bined(_result))
def gcs_matching(L, R, Seeds, auxilary, tau = TAU, mu = MU, show = True): # little tau is first pass screening, ie. "We require a 90% positive match" # mu is an exploration threshold, ie. "We will explore if there is significant improvement" (height, width) = shape = L.shape # ... Tau = Tau_initialize(shape) Tau_count = 0 # Best costs CC = 0-numpy.inf*numpy.ones(shape) CC_prime = 0-numpy.inf*numpy.ones(shape) # Best disparities DD = 0-(1+DISPARITY)*numpy.ones(shape) DD_prime = 0-(1+DISPARITY)*numpy.ones(shape) if show: import matplotlib.pyplot as pyplot figure, axis = pyplot.subplots() pyplot.plot() pyplot.hold(True) while not Seeds.empty(): (_, ss) = Seeds.get() for qq in get_best_neighbours(L, R, ss, auxilary): (vv, uu, uu_prime) = qq cc = get_similarity(L, R, qq, auxilary) #T = 0.9 - float(Tau_count) / (height*width) if (cc < tau): continue if Tau_contains(Tau, qq): continue if (cc+mu < numpy.min([ CC[vv][uu], CC_prime[vv][uu_prime] ])): continue Tau_insert(Tau, qq) Tau_count += 1 Seeds.put( (0-cc, qq) ) if (CC[vv][uu] < cc): CC[vv][uu] = cc DD[vv][uu] = get_disparity(qq) if (CC_prime[vv][uu_prime] < cc): CC_prime[vv][uu_prime] = cc DD_prime[vv][uu_prime] = get_disparity(qq) if show and (Tau_count % 500) == 0: image.show(DD, block = False) figure.canvas.draw() axis.cla() return (DD, DD_prime, CC, CC_prime)
height = numpy.min([L.shape[0], R.shape[0]])
width = numpy.min([L.shape[1], R.shape[1]])
L = L[:height, :width]
R = R[:height, :width]
# TODO: Scale rectify_mask?
# Scale by height
if (MAX_WIDTH < width):
aspect = float(MAX_WIDTH) / width
L = resample(L, [int(aspect * x) for x in L.shape])
R = resample(R, [int(aspect * x) for x in R.shape])
print("Left shape: " + str(L.shape))
print("Right shape: " + str(R.shape))
(L_disparity_map, R_disparity_map, L_cost_map,
R_cost_map) = gcs_matching(L, R)
pickle.dump(L_disparity_map, open(L_disparity_map_path, "wb"))
pickle.dump(R_disparity_map, open(R_disparity_map_path, "wb"))
image.show(L_disparity_map, title="Left Dispartiy Map")
image.show(R_disparity_map, title="Right Dispartiy Map")
image.write("disparity_1.png", L_disparity_map)
image.write("disparity_2.png", R_disparity_map)
