def check_input_file(path, masks):
ret, img_1 = load_img_by_cv2(path + masks[0], grayscale=True)
assert (ret == 0)
height, width = img_1.shape
num_img = len(masks)
for next_index in range(1, num_img):
next_ret, next_img = load_img_by_cv2(path + masks[next_index],
grayscale=True)
assert (next_ret == 0)
next_height, next_width = next_img.shape
assert (height == next_height and width == next_width)
return height, width
def combine_masks(input_dict):
FOREGROUND = input_dict['foreground']
road_mask = input_dict['road_mask']
building_mask = input_dict['building_mask']
output_file = input_dict['save_mask']
files = [road_mask, building_mask]
print("road mask:{}".format(road_mask))
print("building mask:{}".format(building_mask))
height, width = check_input_file(files)
final_mask = np.zeros((height, width), np.uint8)
for idx, file in enumerate(files):
ret, img = load_img_by_cv2(file, grayscale=True)
assert (ret == 0)
label_value = 0
if idx == 0:
label_value = ROAD_VALUE
else:
label_value = BUILDING_VALUE
# label_value = idx+1
# print("buildings prior")
print("Roads prior")
for i in tqdm(range(height)):
for j in range(width):
if img[i, j] >= FOREGROUND:
# print ("img[{},{}]:{}".format(i,j,img[i,j]))
if label_value == ROAD_VALUE:
final_mask[i, j] = label_value
elif label_value == BUILDING_VALUE and final_mask[
i, j] != ROAD_VALUE:
final_mask[i, j] = label_value
# if label_value == BUILDING_VALUE:
# final_mask[i, j] = label_value
# elif label_value == ROAD_VALUE and final_mask[i, j] != BUILDING_VALUE:
# final_mask[i, j] = label_value
final_mask[final_mask == ROAD_VALUE] = 1
final_mask[final_mask == BUILDING_VALUE] = 2
plt.imshow(final_mask, cmap='gray')
plt.title("combined mask")
plt.show()
cv2.imwrite(output_file, final_mask)
print("Saved to : {}".format(output_file))
def vote_masks(input_dict):
target_values = input_dict['target_values']
num_target = len(target_values)
input_files = input_dict['input_files']
output_file = input_dict['save_mask']
files = input_files.split(';')
height, width = check_input_file(files)
mask_list = []
for file in files:
ret, img = load_img_by_cv2(file, grayscale=True)
print(file)
assert (ret == 0)
mask_list.append(img)
vote_mask = np.zeros((height, width), np.uint8)
for i in tqdm(range(height)):
for j in range(width):
# record=np.zeros(256,np.uint8)
record = np.zeros(num_target, np.uint8)
for n in range(len(mask_list)):
mask = mask_list[n]
pixel = mask[i, j]
record[pixel] += 1
# """Alarming"""
if record.argmax(
) == 0: # if argmax of 0 = 125 or 255, not prior considering background(0)
record[0] -= 1
# print("record:{}".format(record))
# a = record[1:]
# print("else:{}".format(a))
# if a.any()>1:
# record[0]=0
label = record.argmax()
# print ("{},{} label={}".format(i,j,label))
vote_mask[i, j] = label
# vote_mask[vote_mask==125]=1
# vote_mask[vote_mask == 255] = 2
print(np.unique(vote_mask))
cv2.imwrite(output_file, vote_mask)
return 0
def combine_all_mask(height, width, input_path, mask_pool):
"""
:param height:
:param width:
:param input_path:
:param mask_pool:
:return: final mask from roads_mask and buildings_mask
prior: road(1)>bulidings(2)
"""
final_mask = np.zeros((height, width), np.uint8)
for idx, file in enumerate(mask_pool):
ret, img = load_img_by_cv2(input_path + file, grayscale=True)
assert (ret == 0)
label_value = 0
if 'road' in file:
label_value = ROAD_VALUE
elif 'building' in file:
label_value = BUILDING_VALUE
# label_value = idx+1
# print("buildings prior")
print("Roads prior")
for i in tqdm(range(height)):
for j in range(width):
if img[i, j] >= FOREGROUND:
# print ("img[{},{}]:{}".format(i,j,img[i,j]))
if label_value == ROAD_VALUE:
final_mask[i, j] = label_value
elif label_value == BUILDING_VALUE and final_mask[
i, j] != ROAD_VALUE:
final_mask[i, j] = label_value
# if label_value == BUILDING_VALUE:
# final_mask[i, j] = label_value
# elif label_value == ROAD_VALUE and final_mask[i, j] != BUILDING_VALUE:
# final_mask[i, j] = label_value
final_mask[final_mask == ROAD_VALUE] = 1
final_mask[final_mask == BUILDING_VALUE] = 2
return final_mask
def vote_per_image(height, width, path, masks):
num_target = len(target_values)
mask_list = []
for tt in range(len(masks)):
ret, img = load_img_by_cv2(path + masks[tt], grayscale=True)
assert (ret == 0)
mask_list.append(img)
vote_mask = np.zeros((height, width), np.uint8)
for i in tqdm(range(height)):
for j in range(width):
# record=np.zeros(256,np.uint8)
record = np.zeros(num_target, np.uint8)
for n in range(len(mask_list)):
mask = mask_list[n]
pixel = mask[i, j]
record[pixel] += 1
# """Alarming"""
if record.argmax(
) == 0: # if argmax of 0 = 125 or 255, not prior considering background(0)
record[0] -= 1
# print("record:{}".format(record))
# a = record[1:]
# print("else:{}".format(a))
# if a.any()>1:
# record[0]=0
label = record.argmax()
# print ("{},{} label={}".format(i,j,label))
vote_mask[i, j] = label
# vote_mask[vote_mask==125]=1
# vote_mask[vote_mask == 255] = 2
print(np.unique(vote_mask))
return vote_mask
from tqdm import tqdm
from ulitities.base_functions import load_img_by_cv2, compare_two_image_size
ref_file = '../../data/test/paper/label/cuiping_label.png'
# 1) jian11_test_label, 2) jiangyou_label, 3) yujiang_test_label,
# 4) cuiping_label, 5) shuangliu_1test_label, 6) tongchuan_test_label
pred_file ='../../data/test/paper/voted/' \
'unet_cuiping_4bands_voted0.png'
output_file = '../../data/test/paper/voted/' \
'unet_cuiping_error_map_7c.png'
if __name__ == '__main__':
print("[INFO]Reading images")
ret, ref_img = load_img_by_cv2(ref_file, grayscale=True)
if ret != 0:
print("Open file failed:{}".format(ref_file))
sys.exit(-1)
ret, pred_img = load_img_by_cv2(pred_file, grayscale=True)
if ret != 0:
print("Open file failed:{}".format(pred_file))
sys.exit(-2)
compare_two_image_size(ref_img, pred_img, grayscale=True)
height, width = ref_img.shape
print("height,width: {},{}".format(height, width))
match_img = np.zeros((height, width), np.uint8)
def accuracy_evalute(input_dict):
ref_file = input_dict['gt_file']
pred_file = input_dict['mask_file']
valid_labels = input_dict['valid_values']
n_class = len(valid_labels)
check_rate = input_dict['check_rate']
gup_id = input_dict['GPUID']
os.environ["CUDA_VISIBLE_DEVICES"] = gup_id
ret, ref_img = load_img_by_cv2(ref_file, grayscale=True)
if ret != 0:
print("Open file failed: {}".format(ref_file))
sys.exit(-1)
ret, pred_img = load_img_by_cv2(pred_file, grayscale=True)
print(np.unique(pred_img))
if ret != 0:
print("Open file failed: {}".format(pred_file))
sys.exit(-2)
print("\nfile: {}".format(os.path.split(pred_file)[1]))
print("[INFO] Calculate confusion matrix..\n")
ref_img = np.array(ref_img)
height, width = ref_img.shape
print(height, width)
if height != pred_img.shape[0] or width != pred_img.shape[1]:
print("image sizes of reference and predicted are not equal!\n")
img_length = height * width
assert (check_rate > 0.001 and check_rate <= 1.00)
num_checkpoints = np.int(img_length * check_rate)
pos = random.sample(range(img_length), num_checkpoints)
"""reshape images from two to one dimension"""
ref_img = np.reshape(ref_img, height * width)
pred_img = np.reshape(pred_img, height * width)
labels = ref_img[pos]
"""ignore nodata pixels"""
valid_index = []
for tt in valid_labels:
ind = np.where(labels == tt)
ind = list(ind)
valid_index.extend(ind[0])
valid_index.sort()
valid_num_checkpoints = len(valid_index)
print("{}points have been selected, but {} points will be used to evaluate accuracy!\n".format(num_checkpoints,
valid_num_checkpoints))
# valid_ref=ref_img[valid_index]
valid_ref = labels[valid_index]
print("valid value in reference image: {}".format(np.unique(valid_ref)))
ts = pred_img[pos]
valid_pred = ts[valid_index]
print("valid value in predicton image: {}".format(np.unique(valid_pred)))
tmp = np.unique(valid_pred)
for ss in tmp:
assert (ss in valid_labels)
"""Test to find out where are labels in the confusion matrix"""
# num_tmp = labels[labels==0].shape
# print("label 0 : {}".format(num_tmp))
# num_tmp = labels[labels == 1].shape
# print("label 1 : {}".format(num_tmp))
# num_tmp = labels[labels == 2].shape
# print("label 2 : {}".format(num_tmp))
# confus_matrix = tf.contrib.metrics.confusion_matrix(labels, predictions, n_class)
confus_matrix = tf.contrib.metrics.confusion_matrix(valid_pred, valid_ref, n_class)
with tf.Session() as sess:
confus_matrix = sess.run(confus_matrix)
print(confus_matrix)
confus_matrix = np.array(confus_matrix)
oa = 0
x_row_plus = []
x_col_plus = []
x_diagonal = []
for i in range(n_class):
oa += confus_matrix[i, i]
x_diagonal.append(confus_matrix[i, i])
row_sum = sum(confus_matrix[i, :])
col_sum = sum(confus_matrix[:, i])
x_row_plus.append(row_sum)
x_col_plus.append(col_sum)
print("x_row_plus:{}".format(x_row_plus))
print("x_col_plus:{}".format(x_col_plus))
x_row_plus = np.array(x_row_plus)
x_col_plus = np.array(x_col_plus)
x_diagonal = np.array(x_diagonal)
x_total = sum(x_row_plus)
OA_acc = oa / (sum(x_row_plus))
print("\nOA:{:.3f}".format(OA_acc))
tmp = x_col_plus * x_row_plus
kappa = (x_total * sum(x_diagonal) - sum(x_col_plus * x_row_plus)) / np.float(
x_total * x_total - sum(x_col_plus * x_row_plus))
print("Kappa:{:.3f}".format(kappa))
for i in range(n_class - 1):
i = i + 1
prec = x_diagonal[i] / x_row_plus[i]
print("\nForground of {}_accuracy= {:.3f}".format(i, prec))
recall = x_diagonal[i] / x_col_plus[i]
print("{}_recall= {:.3f}".format(i, recall))
iou = x_diagonal[i] / (x_row_plus[i] + x_col_plus[i] - x_diagonal[i])
print("{}_iou {:.3f}".format(i, iou))
files,num = get_file(input_label_path)
assert (num!=0)
# valid_labels = []
# if FLAG_USING_UNET:
# valid_labels = unet_labels
# else:
# valid_labels = segnet_labels
for label_file in tqdm(files):
# label_file = input_label_path + os.path.split(src_file)[1]
#
# ret,src_img = load_img(src_file)
# assert(ret==0)
ret,label_img = load_img_by_cv2(label_file, grayscale=True)
assert (ret == 0)
local_labels = np.unique(label_img)
invalid_labels=[]
for tmp in local_labels:
if tmp not in valid_labels:
invalid_labels.append(tmp)
print ("\nWarning: some label is not valid value")
print ("\nFile: {}".format(label_file))
HAS_INVALID_VALUE = True
if HAS_INVALID_VALUE == True:
new_label_img = make_label_valid(label_img, invalid_labels)
