def crop_by_mask(img_ori,
img_mask,
padding_size=20,
seg_padding_size=0,
rotate=True):
# add pad to seg image
if seg_padding_size > 0:
_, contours, hierachy = cv2.findContours(img_mask, cv2.RETR_EXTERNAL,
cv2.CHAIN_APPROX_NONE)
img_mask = cv2.drawContours(img_mask, contours, -1, 255,
seg_padding_size)
# rotate image
if rotate:
angle = calculate_angle(img_mask)
img_mask = rotate_bound(img_mask, angle)
img_ori = rotate_bound(img_ori, angle)
# Add black padding to input image and seg image
img_mask = cv2.copyMakeBorder(img_mask, padding_size, padding_size,
padding_size, padding_size, 0)
img_ori = cv2.copyMakeBorder(img_ori, padding_size, padding_size,
padding_size, padding_size, 0)
# get white pixel bounding box
x, y, w, h = find_bounding_square(img_mask, padding=padding_size)
img_crop = img_ori[int(y):int(y + h), int(x):int(x + w)]
return img_crop
def main(seg_path, us_path, mask_path, kidney_list_csv, save_path, padding_size=20, train=True):
seg_path = norm_path(seg_path)
us_path = norm_path(us_path)
mask_path = norm_path(mask_path)
save_path = norm_path(save_path)
#csv 파일 경로 추가
kidney_list_csv = norm_path(kidney_list_csv)
if not os.path.exists(save_path):
os.makedirs(os.path.join(save_path, 'kidney'))
os.makedirs(os.path.join(save_path, 'non-kidney'))
seg_name = file_dict(seg_path)
us_name = file_dict(us_path)
mask_name = file_dict(mask_path)
kidney_set = kidney_file_set(kidney_list_csv)
for name_order in tqdm(mask_name.keys()):
name, order = name_order.split('#')
mask_img = cv2.imread(mask_name[name_order], cv2.IMREAD_GRAYSCALE)
us_img = cv2.imread(us_name[name], cv2.IMREAD_GRAYSCALE)
# Add black padding to input image and seg image
mask_img = cv2.copyMakeBorder(mask_img, padding_size, padding_size, padding_size, padding_size, 0)
us_img = cv2.copyMakeBorder(us_img, padding_size, padding_size, padding_size, padding_size, 0)
# rotate image
angle = calculate_angle(mask_img)
mask_img = rotate_bound(mask_img, angle)
us_img = rotate_bound(us_img, angle)
# get white pixel bounding box
x, y, w, h = find_bounding_square(mask_img, padding=padding_size)
# crop bounding
mask_img = mask_img[int(y):int(y + h), int(x):int(x + w)]
us_img = us_img[int(y):int(y + h), int(x):int(x + w)]
# mask image
shape_img = np.where(mask_img == 255, us_img, 0)
# has kidney?
if order == '0':
if train:
has_kidney = name in seg_name.keys()
else:
has_kidney = name in kidney_set
else:
has_kidney = False
dst_path = os.path.join(save_path, 'kidney' if has_kidney else 'non-kidney')
dst_file = os.path.join(dst_path, name_order + '.png')
if not os.path.exists(dst_path):
os.makedirs(dst_path)
# print(dst_file)
cv2.imwrite(dst_file, shape_img)
def image_crop_kidney_loader(input_path, seg_path):
# input_path : non-kidney image
# seg_path : any seg image
# output : crop non-kidney image
padding_size = args.kidney_crop_padding
# read image in grayscale
input_image = cv2.imread(input_path, cv2.IMREAD_GRAYSCALE)
seg_image = cv2.imread(seg_path, cv2.IMREAD_GRAYSCALE)
seg_image = cv2.resize(seg_image,
(input_image.shape[1], input_image.shape[0]))
# rotate image
angle = calculate_angle(seg_image)
seg_image = rotate_bound(seg_image, angle)
input_image = rotate_bound(input_image, angle)
# Add black padding to input image and seg image
seg_image = cv2.copyMakeBorder(seg_image, padding_size, padding_size,
padding_size, padding_size, 0)
input_image = cv2.copyMakeBorder(input_image, padding_size, padding_size,
padding_size, padding_size, 0)
# get white pixel bounding box
x, y, w, h = find_bounding_square(seg_image, padding=padding_size)
# crop kidney with padding
result_image = input_image[int(y):int(y + h), int(x):int(x + w)]
# cv2 img into pil img
img_for_pil = Image.fromarray(result_image)
resized_img = pil_resize(img_for_pil)
# #debug
# debug_path = os.path.join(args.result, 'debug')
# if not os.path.exists(debug_path):
# os.makedirs(debug_path)
# debug_file = os.path.join(debug_path, os.path.split(input_path)[1])
# cv2.imwrite(debug_file, result_image)
return resized_img
def main(us_path,
mask_path,
result_path,
scale=None,
view_scale=None,
use_top_n_greatest_width=True,
size=None,
horizontally=True,
crop=False,
denoise=True):
'''
center boolean: 신장의 중앙 정렬 여부
scale float: 영상 축적 고정
only_greatest_width boolean: 신장 영상중 가장 큰것만 사용할지 여부
resize (W, H): 최종 이미지 크기
'''
us_dict = image_dict(norm_path(us_path))
mask_dict = image_dict(norm_path(mask_path))
result_path = norm_path(result_path, True)
for patient, dicoms in Per_patient_fast():
patient_datas = list()
for dicom in dicoms:
Name = dicom['Name'][:-4] # remove file ext
Diagnosis = dicom['Diagnosis']
AccNo = dicom['AccNo']
PhysicalUnitsXDirection = dicom['PhysicalUnitsXDirection']
PhysicalUnitsYDirection = dicom['PhysicalUnitsYDirection']
PhysicalDeltaY = dicom['PhysicalDeltaY'] # cm per pixels
PhysicalDeltaX = dicom['PhysicalDeltaX'] # 1 픽셀이 몇 cm 인지 나타냄
# checking the kidney
if Name not in mask_dict:
# pass the non-kidney images
continue
# for debug
# if Name != '1.2.840.113663.1500.1.295077244.3.8.20101014.92756.625':
# continue
# checking physical unit(mm)
if PhysicalUnitsXDirection != '3' or PhysicalUnitsYDirection != '3' \
and not PhysicalDeltaY and not PhysicalDeltaX:
continue
PhysicalDeltaY = float(PhysicalDeltaY)
PhysicalDeltaX = float(PhysicalDeltaX)
# read image
us_img = cv2.imread(us_dict[Name], cv2.IMREAD_GRAYSCALE)
print(us_dict[Name])
mask_img = cv2.imread(mask_dict[Name], cv2.IMREAD_GRAYSCALE)
# assert scale != view_scale, "use only scale or view_scale"
if horizontally:
angle = calculate_angle(mask_img)
mask_img = rotate_bound(mask_img, angle)
us_img = rotate_bound(us_img, angle)
if scale:
us_img = resize_physical_unit(us_img,
(PhysicalDeltaX, PhysicalDeltaY),
scale)
mask_img = resize_physical_unit(
mask_img, (PhysicalDeltaX, PhysicalDeltaY), scale)
if view_scale:
bbX, bbY, bbW, bbH = find_bounding_square(mask_img)
fx = view_scale[0] / bbW
fy = view_scale[1] / bbH
us_img = cv2.resize(us_img, None, fx=fx, fy=fy)
mask_img = cv2.resize(mask_img, None, fx=fx, fy=fy)
if size:
#numpy 좌표값을 얻어낸다 -> argwhere
mask_pts = np.argwhere(mask_img == 255)
#평균을 구한다
mask_cx, mask_cy = int(np.mean(mask_pts[:, 1])), int(
np.mean(mask_pts[:, 0]))
#캔버스를 만들고 붙혀넣는다
us_canvas = np.zeros(size, dtype=np.uint8)
mask_canvas = np.zeros(size, dtype=np.uint8)
# # for debug
# us_canvas[:] = 100
# mask_canvas[:] = 100
canvas_h, canvas_w = size
canvas_cx, canvas_cy = canvas_w // 2, canvas_h // 2
dst_x, dst_y = canvas_cx - mask_cx, canvas_cy - mask_cy
us_img = paste(dst=us_canvas,
src=us_img,
dst_x=dst_x,
dst_y=dst_y)
mask_img = paste(dst=mask_canvas,
src=mask_img,
dst_x=dst_x,
dst_y=dst_y)
if crop:
mask = mask_img.astype(np.bool)
us_img = us_img * mask
if denoise:
us_img = cv2.fastNlMeansDenoising(us_img, None, 10, 7, 21)
data = dict()
data['us'] = us_img
data['mask'] = mask_img
data['info'] = dicom
patient_datas.append(data)
# sort by kidney value
patient_datas = sorted(
patient_datas,
key=lambda x: np.count_nonzero(x['mask'] == 255),
reverse=True)
# use one or all
if use_top_n_greatest_width:
cut_n = use_top_n_greatest_width
patient_datas = patient_datas[:cut_n]
accNo = patient['AccNo']
diagnosis = patient['Diagnosis']
train_val = 'val' if accNo in isangmi_accno_list else 'train'
save_path = os.path.join(result_path, train_val, diagnosis, accNo)
''' if args.preprocess_denoise:
np_out = np.asarray(out)
np_out = cv2.fastNlMeansDenoising(np_out, None, 10, 7, 21)
out = Image.fromarray(np.uint8(np_out))
return out'''
# save image
for order, data in enumerate(patient_datas):
us_img = data['us']
name = data['info']['Name'][:-4]
save_file = os.path.join(save_path, name + '.png')
if not os.path.exists(save_path):
os.makedirs(save_path)
print(save_file)
cv2.imwrite(save_file, us_img)
# save order of images
if patient_datas:
save_file = os.path.join(save_path, 'order.txt')
with open(save_file, 'wt') as f:
for order, data in enumerate(patient_datas):
name = data['info']['Name'][:-4]
f.write(name + '\n')
def main(us_path, mask_path, save_path, padding_size=20, train=True):
us_path = norm_path(us_path)
mask_path = norm_path(mask_path)
save_path = norm_path(save_path)
if not os.path.exists(save_path):
os.makedirs(os.path.join(save_path, 'normal'))
os.makedirs(os.path.join(save_path, 'AKI'))
os.makedirs(os.path.join(save_path, 'CKD'))
us_name = file_dict(us_path)
mask_name = file_dict(mask_path)
for name_order in tqdm(mask_name.keys()):
print('name_order:', name_order)
if '#' in name_order:
name, order = name_order.split('#')
print('name:', name)
print('order:', order)
else:
name = name_order
continue
print('no has: #')
mask_img = cv2.imread(mask_name[name_order], cv2.IMREAD_GRAYSCALE)
us_img = cv2.imread(us_name[name], cv2.IMREAD_GRAYSCALE)
# Add black padding to input image and seg image
mask_img = cv2.copyMakeBorder(mask_img, padding_size, padding_size,
padding_size, padding_size, 0)
us_img = cv2.copyMakeBorder(us_img, padding_size, padding_size,
padding_size, padding_size, 0)
# rotate image
angle = calculate_angle(mask_img)
mask_img = rotate_bound(mask_img, angle)
us_img = rotate_bound(us_img, angle)
# get white pixel bounding box
x, y, w, h = find_bounding_square(mask_img, padding=padding_size)
# crop bounding
mask_img = mask_img[int(y):int(y + h), int(x):int(x + w)]
us_img = us_img[int(y):int(y + h), int(x):int(x + w)]
# mask image
shape_img = np.where(mask_img == 255, us_img, 0)
result_img = preproess_image(shape_img, mask_img)
# has kidney?
if order == '0':
if train:
has_kidney = name in seg_name.keys()
else:
has_kidney = name in kidney_set
else:
has_kidney = False
dst_path = os.path.join(save_path,
'kidney' if has_kidney else 'non-kidney')
dst_file = os.path.join(dst_path, name_order + '.png')
if not os.path.exists(dst_path):
os.makedirs(dst_path)
# print(dst_file)
cv2.imwrite(dst_file, result_img)
def main(us_path, mask_path, save_path, padding_size=20):
us_path = norm_path(us_path)
mask_path = norm_path(mask_path)
save_path = norm_path(save_path)
if not os.path.exists(save_path):
os.makedirs(os.path.join(save_path, 'normal'))
os.makedirs(os.path.join(save_path, 'AKI'))
os.makedirs(os.path.join(save_path, 'CKD'))
us_name = file_dict(us_path)
mask_name = file_dict(mask_path)
for name_order in tqdm(mask_name.keys()):
if '#' in name_order:
name, order = name_order.split('#')
print('saved:', name)
if name not in mask_name.keys():
print('no:', name)
continue
else:
print('no has: #')
continue
mask_img = cv2.imread(mask_name[name_order], cv2.IMREAD_GRAYSCALE)
us_img = cv2.imread(us_name[name], cv2.IMREAD_GRAYSCALE)
# Add black padding to input image and seg image
mask_img = cv2.copyMakeBorder(mask_img, padding_size, padding_size,
padding_size, padding_size, 0)
us_img = cv2.copyMakeBorder(us_img, padding_size, padding_size,
padding_size, padding_size, 0)
# rotate image
angle = calculate_angle(mask_img)
mask_img = rotate_bound(mask_img, angle)
us_img = rotate_bound(us_img, angle)
# get white pixel bounding box
x, y, w, h = find_bounding_square(mask_img, padding=padding_size)
# crop bounding
mask_img = mask_img[int(y):int(y + h), int(x):int(x + w)]
us_img = us_img[int(y):int(y + h), int(x):int(x + w)]
# mask image
shape_img = np.where(mask_img == 255, us_img, 0)
result_img = preproess_image(shape_img, mask_img)
KidneyList = ['normal', 'AKI', "CKD"]
if KidneyList[0] in us_name[name]:
dst_path = os.path.join(save_path, KidneyList[0])
print('save normal')
elif KidneyList[1] in us_name[name]:
dst_path = os.path.join(save_path, KidneyList[1])
print('save AKI')
elif KidneyList[2] in us_name[name]:
dst_path = os.path.join(save_path, KidneyList[2])
print('save CKD')
else:
continue
dst_file = os.path.join(dst_path, name_order + '.png')
if not os.path.exists(dst_path):
os.makedirs(dst_path)
# print(dst_file)
cv2.imwrite(dst_file, result_img)