def QuoteAndImage():
url = 'https://www.channel24.co.za/ShowMax/10-unforgettable-lines-from-friends-20161109'
soup = f.request(url)
quotes = soup.select("strong")
quotes = [q.text for q in quotes][0:-1]
quotes.pop(8)
quotes.pop(9)
quotes.pop(7)
quotes_characters = {}
for q in quotes:
n = q.split('.')[1]
quotes_characters[n.split(':')[0]] = n.split(':')[1]
for i in [' Joey', ' Rachel', ' Phoebe']:
quotes_characters[i] = quotes_characters[i].replace('’', "'")
images = soup.select("div[class='embed image'] img[src]")
images.pop(2)
images.pop(-1)
images.pop(3)
images.pop(-2)
images = [images[i]['src'] for i in range(len(images))]
character = ["Joey", "Monica", "Ross", "Rachel", "Phoebe", "Chandler"]
i = 0
for image in images:
f.save_image('OUTPUT/' + character[i] + '.jpg', image)
i += 1
url = 'https://es.wikipedia.org/wiki/Archivo:Friends_logo.svg'
soup = f.request(url)
logo = 'https:' + soup.select(
'#file > a:nth-child(1) > img:nth-child(1)')[0]['src']
f.save_image('OUTPUT/logo.svg.png', logo)
return quotes_characters
def main(options):
#This program capture one image each 1 second
cam=camera.camera(options.wt,options.ht)
cam.capture_image()
anterior_image=cam.actual_image
while True:
cam.capture_image()
cam.save_image('./prof1.png')
fs.save_image('./prof2.png',anterior_image)
anterior_image=cam.actual_image
imr=fs.differences_images('./prof1.png','./prof2.png')
fs.save_cv2_image('./result.png',imr)
imr = pygame.image.load('./result.png')
cam.image_show=imr
cam.show_image()
#cam.show_image()
cam.delay_camera(options.time)
evaluator = Evaluator()
def main():
pass
if __name__ == "__main__":
main()
x = functions.preprocess_image(content_image_path, width, height)
for i in range(iterations):
print('Start of iteration', i)
start_time = time.time()
x, min_val, info = fmin_l_bfgs_b(evaluator.loss,
x.flatten(),
fprime=evaluator.grads,
maxfun=20)
print('Current loss value:', min_val)
# save current generated image
img = functions.deprocess_image(x.copy(), img_nrows, img_ncols)
fname = result_prefix + '_at_iteration_%d.png' % i
functions.save_image(img, fname)
# image.save_img(fname, img)
end_time = time.time()
print('Image saved as', fname)
print('Iteration %d completed in %ds' % (i, end_time - start_time))
def style_transfer(content_image_path, style_image_path, result_prefix='transfer_', iterations=1, content_weight=0.025, style_weight = 1.0, total_variation_weight = 1):
# dimensions of the generated picture.
evaluator = Evaluator()
content_weight = float(content_weight)
style_weight = float(style_weight)
result_prefix = TRANSFER_FOLDER + "/" + str(int(time.time()))+ "_" + result_prefix
variables.width, variables.height = functions.load_image(content_image_path).size
content_image = K.variable(functions.preprocess_image(content_image_path, variables.width, variables.height))
style_reference_image = K.variable(functions.preprocess_image(style_image_path,variables.width, variables.height))
img_nrows, img_ncols = functions.calc_rowsandcols(variables.width,variables.height)
if K.image_data_format() == 'channels_first':
combination_image = K.placeholder((1, 3, img_nrows, img_ncols))
else:
combination_image = K.placeholder((1, img_nrows, img_ncols, 3))
# print_all(content_image)
# print_all(style_reference_image)
# print_all(combination_image)
input_tensor = K.concatenate([content_image,
style_reference_image,
combination_image], axis=0)
# print_all(input_tensor)
# exit(0)
model = functions.customVGG16(input_tensor)
outputs_dict = dict([(layer.name, layer.output) for layer in model.layers])
loss = K.variable(0.)
layer_features = outputs_dict['conv2d_12']
content_image_features = layer_features[0, :, :, :]
combination_features = layer_features[2, :, :, :]
loss += content_weight * functions.content_loss(content_image_features,
combination_features)
feature_layers = ['conv2d_1', 'conv2d_3',
'conv2d_5', 'conv2d_8',
'conv2d_11']
for layer_name in feature_layers:
layer_features = outputs_dict[layer_name]
style_reference_features = layer_features[1, :, :, :]
combination_features = layer_features[2, :, :, :]
sl = functions.style_loss(style_reference_features, combination_features,variables.width,variables.height)
loss += (style_weight / len(feature_layers)) * sl
loss += total_variation_weight * functions.total_variation_loss(combination_image,variables.width,variables.height)
grads = K.gradients(loss, combination_image)
outputs = [loss]
if isinstance(grads, (list, tuple)):
outputs += grads
else:
outputs.append(grads)
variables.f_outputs = K.function([combination_image], outputs)
x = functions.preprocess_image(content_image_path,variables.width,variables.height)
start_time = time.time()
for i in range(int(iterations)):
print('Start of iteration', i)
iter_start_time = time.time()
x, min_val, info = fmin_l_bfgs_b(evaluator.loss, x.flatten(),
fprime=evaluator.grads, maxfun=20, maxiter=1)
print('Current loss value:', min_val)
end_time = time.time()
print('Iteration %d completed in %ds' % (i, end_time - iter_start_time))
# return [min_val,fname,i,end_time-start_time]
# save current generated image
img = functions.deprocess_image(x.copy(),img_nrows, img_ncols)
fname = result_prefix + 'at_iteration_%d.png' % i
functions.save_image(img,fname)
values = [iterations,str(round(min_val,2)),str(round(end_time-start_time,2)),fname]
return values
# VOIs = ['Lobule II', 'Lobules IV-V', 'Substantia nigra, compact part', 'Substantia nigra, reticular part']
# VOIs = ['Substantia nigra, compact part', 'Substantia nigra, reticular part']
# Load
structure = pd.read_csv(reference_structure_path)
volume_table = pd.read_csv(analysis_table_path)
# Convert reference structure table
id_custom_to_id_mc = [762, 500, 821, 300, 977, 350]
id_mc_to_id_custom = [500, 762, 300, 821, 350, 977]
structure['id_mc'] = npi.remap(structure['id_custom'],
id_custom_to_id_mc,
id_mc_to_id_custom)
structure.to_csv(reference_structure_mc_path)
# Convert reference and data annotation files
for Path in mouse_path_list + [annotation_path] + mouse_lobular_path_list:
print(Path)
annotation_image = nib.load(Path)
annotation = annotation_image.get_fdata()
annotation = remap_3D(annotation, id_custom_to_id_mc, id_mc_to_id_custom)
print(Path)
output_path = Path.split('.')[0]+'_mc.nii.gz'
print(output_path)
save_image(annotation, annotation_image, output_path)
def open_file():
#clear global list of indices
for i in img_idx:
img_idx.pop()
img_idx.append(0) #set global index to 0
browse_text.set("loading...")
#load a PDF file
file = askopenfile(parent=root,
mode='rb',
filetypes=[("Pdf file", "*.pdf")])
if file:
read_pdf = PyPDF2.PdfFileReader(file)
#select a page
page = read_pdf.getPage(0)
#extract text content from page
page_content = page.extractText()
#SET A SPECIAL ENCODING OR REPLACE CHARACTERS
#page_content = page_content.encode('cp1252')
page_content = page_content.replace('\u2122', "'")
#CLEARING GLOBAL VARIABLES ONCE A NEW PDF FILE IS SELECTED
#clear the content of the previous PDF file
if all_content:
for i in all_content:
all_content.pop()
#clear the image list from the previous PDF file
for i in range(0, len(all_images)):
all_images.pop()
#hide the displayed image from the previous PDF file and remove it
if displayed_img:
displayed_img[-1].grid_forget()
displayed_img.pop()
#BEGIN EXTRACTING
#extract text
all_content.append(page_content)
#extract images
images = extract_images(page)
for img in images:
all_images.append(img)
#BEGIN DISPLAYING
#display the first image that was detected
selected_image = display_images(images[img_idx[-1]])
displayed_img.append(selected_image)
#display the text found on the page
display_textbox(all_content, 4, 0, root)
#reset the button text back to Browse
browse_text.set("Browse")
#BEGIN MENUES AND MENU WIDGETS
#1.image menu on row 2
img_menu = Frame(root, width=800, height=60)
img_menu.grid(columnspan=3, rowspan=1, row=2)
what_text = StringVar()
what_img = Label(root, textvariable=what_text, font=("shanti", 10))
what_text.set("image " + str(img_idx[-1] + 1) + " out of " +
str(len(all_images)))
what_img.grid(row=2, column=1)
#arrow buttons
display_icon('arrow_l.png', 2, 0, E,
lambda: left_arrow(all_images, selected_image, what_text))
display_icon(
'arrow_r.png', 2, 2, W,
lambda: right_arrow(all_images, selected_image, what_text))
#2.save image menu on row 3
save_img_menu = Frame(root, width=800, height=60, bg="#c8c8c8")
save_img_menu.grid(columnspan=3, rowspan=1, row=3)
#create action buttons
copyText_btn = Button(root,
text="copy text",
command=lambda: copy_text(all_content, root),
font=("shanti", 10),
height=1,
width=15)
saveAll_btn = Button(root,
text="save all images",
command=lambda: save_all(all_images),
font=("shanti", 10),
height=1,
width=15)
save_btn = Button(root,
text="save image",
command=lambda: save_image(all_images[img_idx[-1]]),
font=("shanti", 10),
height=1,
width=15)
#place buttons on grid
copyText_btn.grid(row=3, column=0)
saveAll_btn.grid(row=3, column=1)
save_btn.grid(row=3, column=2)
def test_compare_darken():
out = do_darken(test_image, args_mock)
save_image(out, "test_images/darkenOut.png")
output = read_image("test_images/darkenOut.png")
test_input = read_image("test_images/darken.png")
assert (output == test_input).all()
def test_compare_bw():
out = do_bw(test_image)
save_image(out, "test_images/bwOut.png")
output = read_image("test_images/bwOut.png")
test_input = read_image("test_images/bw.png")
assert (output == test_input).all()
# Extract cerebellum and substantia nigra for references
for i in range(len(annotation_path_list)):
annotation_image = nib.load(annotation_path_list[i])
annotation = annotation_image.get_fdata().astype(int)
annotation = annotation.astype(int)
template_image = nib.load(template_path_list[i])
template = template_image.get_fdata()
annotation_in_structure = np.isin(annotation, ids_list[i])
annotation_structure = annotation * annotation_in_structure
template_structure = template * annotation_in_structure
save_image(annotation_structure, annotation_image, annotation_path_list[i].split('.')[0] + '_' + structure_name_list[i] + '.nii.gz')
save_image(template_structure, template_image, template_path_list[i].split('.')[0] + '_' + structure_name_list[i] + '.nii.gz')
# For each subject create cerebellum isolated files for both inmasked template and annotation
annotation_path_list_list = [glob.glob(os.path.join(data_path, '*', '*orsuit_thrarg*lobular_mc.nii.gz')),
glob.glob(os.path.join(data_path, '*', '*subcortical_thrarg.nii.gz'))]
template_path_list = glob.glob(os.path.join(data_path, '*', '*reoriented.nii.gz'))
for i in range(len(annotation_path_list_list)):
annotation_path_list = annotation_path_list_list[i]
for iSubject in range(len(annotation_path_list)):
annotation_path = annotation_path_list[iSubject]
template_path = template_path_list[iSubject]
annotation_image = nib.load(annotation_path)
folder_path = input_folder_path+'/'+chapter_info
folder_status = os.path.exists(folder_path)
if not folder_status :
os.makedirs(folder_path) #创建每一话文件夹的路径，已存在的话跳过
pages = len(page_url.keys()) #用于提示进度
print('目前正在下载 '+chapter_info+'----------------------------------')
image_list = os.listdir(folder_path+'/')
image_total_num = len(image_list) #确定该文件下已有多少张图片
print('进度-------------------------'+' '+str(image_total_num)+' / '+str(pages))
for page,image_url in page_url.items():
try:
image_path = folder_path+'/'+chapter_info+'-'+str(page)+'.jpg'
image_status = os.path.exists(image_path)
if not image_status: #检测图片是否已下载，未下载的话会进行下载，否则会跳过
image = functions.get_html_resource(image_url,download_info.headers)
functions.save_image(image_path,image)
else:
print(chapter_info+'-'+str(page)+'.jpg'+' 已下载')
continue
except:
print('\n异常发生，暂停10秒后继续>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>')
print('此异常会导致有一张图片的下载被跳过，重新运行此程序后便可补上这张图片>>>>>>>>>>>>>>>>>>>>>\n')
time.sleep(10)
continue
else:
print(chapter_info+'-'+str(page)+'.jpg'+' 下载完成')
time.sleep(random.uniform(0.7,2))
print(chapter_info+' 已完成下载')
def open_file():
for i in img_idx:
img_idx.pop()
img_idx.append(0)
browse_text.set("cargando....")
file = askopenfile(parent=root,
mode="rb",
filetypes=[("PDF file", "*.pdf")])
if file:
read_pdf = PyPDF2.PdfFileReader(file)
page = read_pdf.getPage(0)
page_content = page.extractText()
page_content = page_content.replace('\u2122', "'")
if all_content:
for i in all_content:
all_content.pop()
for i in range(0, len(all_images)):
all_images.pop()
if displayed_img:
displayed_img[-1].grid_forget()
displayed_img.pop()
all_content.append(page_content)
images = extract_images(page)
for img in images:
all_images.append(img)
selected_image = display_images(images[img_idx[-1]])
displayed_img.append(selected_image)
display_textbox(page_content, 4, 0, root)
browse_text.set("Navegar")
img_menu = Frame(root, width=800, height=60)
img_menu.grid(columnspan=3, rowspan=1, row=2)
que_text = StringVar()
que_img = Label(root, textvariable=que_text)
que_text.set("Imagen " + str(img_idx[-1] + 1) + " de " +
str(len(all_images)))
que_img.grid(row=2, column=1)
display_icon('img/arrow_l.png', 2, 0, E,
lambda: left_arrow(all_images, selected_image, que_text))
display_icon('img/arrow_r.png', 2, 2, W,
lambda: right_arrow(all_images, selected_image, que_text))
save_img = Frame(root, width=800, height=60, bg="#c8c8c8")
save_img.grid(columnspan=3, rowspan=1, row=3)
copyText_btn = Button(root,
text="Copiar texto",
command=lambda: copy_text(all_content),
height=1,
width=15)
saveAll_btn = Button(root,
text="Todas las imagenes",
command=lambda: save_all(all_images),
height=1,
width=15)
save_btn = Button(root,
text="Guardar imagen",
command=lambda: save_image(all_images[img_idx[-1]]),
height=1,
width=15)
copyText_btn.grid(row=3, column=0)
saveAll_btn.grid(row=3, column=1)
save_btn.grid(row=3, column=2)
# .merge(categories, on="CategoryId")["CategoryName"].tolist())
target_classes_names = []
for i in target_classes:
target_classes_names.append(categories['CategoryName'][i - 1])
print("Here's an example of one of the images in the development set")
show_image(images[0])
print(true_classes_names[0].split(',')[0].replace(' ', '_'))
#print(categories['CategoryName'][0])
# Saves the image in order for analysis and spit out their correct category
for i in range(len(images)):
save_image(
images[i], i,
'C:/Users/Bowen/Desktop/Project/image-perturbation-defense/NIPS/test/')
correct_label = true_classes_names[i].split(',')[0].replace(
' ', '_') # Convert ['giant panda, panda etc....'] to 'giant_panda'
with open('true_classes.csv', 'a', newline='') as csvfile:
spamwriter = csv.writer(csvfile)
spamwriter.writerow([correct_label])
slim = tf.contrib.slim
tf.logging.set_verbosity(tf.logging.INFO)
batch_shape = [32, image_height, image_width, 3]
image_iterator = load_images(input_dir, batch_shape)
# Purturbs the image in a batches of 32 to avoid memory issue
for x in batch(range(0, 1000), 32):
print(x)
image = img_as_float(
imread('E:\\User\\Desktop\\Khlamskaya_prog\\IPTI\\' + CONST.name + '.jpg'))
w, h, d = image.shape
w = (w // CONST.patch_size) * CONST.patch_size
h = (h // CONST.patch_size) * CONST.patch_size
image = image[0:w, 0:h, :]
# 2. Создайте матрицу объекты-признаки: характеризуйте каждый пиксель тремя координатами - значениями интенсивности
# в пространстве RGB.
pixels = pandas.DataFrame(np.reshape(image, (w * h, d)),
columns=['R', 'G', 'B'])
_w = w // CONST.patch_size
_h = h // CONST.patch_size
radiance = func.get_radiance(image, Const)
#save_radianced_image(radiance, CONST)
#build_histogram(radiance, CONST)
a = func.get_atmosphere_A(image, radiance, CONST)
w, d = radiance.shape
r = radiance.reshape(w * d, 1)
r = np.hstack((r, r, r))
r = r.reshape(w, d, 3)
func.save_image(r, CONST)
'''
for i in range(1,10):
im = boxfilter(np.average(image, axis=2), i)
imsave('E:\\User\\Desktop\\Khlamskaya_prog\\IPTI\\___boxfilter_' + str(i) + '.jpg', im)
'''
# Extract cerebellum for allen atlas allen_image = nib.load(allen_image_path) allen = allen_image.get_fdata() allen = np.round(allen).astype(int) allen_template_image = nib.load(allen_template_image_path) allen_template = allen_template_image.get_fdata() allen_in_cerebellum = np.isin(allen, cerebellum_ids) # cerebellum_voxel_number = np.sum(allen_in_cerebellum) # cerebellum_volume = cerebellum_voxel_number * voxel_reference_volume allen_cerebellum = allen * allen_in_cerebellum allen_template_cerebellum = allen_template * allen_in_cerebellum save_image(allen_cerebellum, allen_image, os.path.join(reference_path, 'annotation_50_reoriented_mc_ci.nii.gz')) save_image(allen_template_cerebellum, allen_template_image, os.path.join(reference_path, 'average_template_50_reoriented_ci.nii.gz')) allen_in_sn = np.isin(allen, sn_ids) allen_sn = allen * allen_in_sn allen_template_sn = allen_template * allen_in_sn save_image(allen_sn, allen_image, os.path.join(reference_path, 'annotation_50_reoriented_mc_si.nii.gz')) save_image(allen_template_sn, allen_template_image, os.path.join(reference_path, 'average_template_50_reoriented_si.nii.gz')) # For each subject create cerebellum isolated files for both inmasked template and annotation input_list = glob.glob(os.path.join(data_path, '*')) # input_list = ['Data\\Mouse\\Processed_Old\\WT_50', 'Data\\Mouse\\Processed_Old\\KO_6',
parser.add_argument("INPUT_FILE",
help="Vstupni soubor na upravu (cestu k nemu)")
parser.add_argument("OUTPUT_FILE", help="Cesta k vystupu tohoto programu")
"""
------------
"""
queue = sys.argv[1:-2] # slice pro poradi prepinacu
args = parser.parse_args()
np_image = None
try:
np_image = read_image(args.INPUT_FILE)
except FileNotFoundError:
print("Soubor nenalezen, ukoncuji")
exit(1)
except Exception as ex:
print("chyba", ex)
exit(1)
for act in queue:
try:
np_image = action_dict[act](np_image, args)
except KeyError: # ciselne a nevalidni hodnoty muzeme preskocit
pass
except Exception as ex:
print("chyba", ex)
save_image(np_image, args.OUTPUT_FILE) # v pripade potreby soubor prepiseme
