def cvt_det(gt_path, save_path, img_folder):
"""
将icdar2015格式的gt转换为json格式
:param gt_path:
:param save_path:
:return:
"""
gt_dict = {'data_root': img_folder}
data_list = []
origin_gt = load(gt_path)
for img_name, gt in tqdm(origin_gt.items()):
cur_gt = {'img_name': img_name + '.jpg', 'annotations': []}
for line in gt:
cur_line_gt = {'polygon': [], 'text': '', 'illegibility': False, 'language': 'Latin'}
chars_gt = [{'polygon': [], 'char': '', 'illegibility': False, 'language': 'Latin'}]
cur_line_gt['chars'] = chars_gt
# 字符串级别的信息
cur_line_gt['polygon'] = line['points']
cur_line_gt['text'] = line['transcription']
cur_line_gt['illegibility'] = line['illegibility']
cur_line_gt['language'] = line['language']
cur_gt['annotations'].append(cur_line_gt)
data_list.append(cur_gt)
gt_dict['data_list'] = data_list
save(gt_dict, save_path)
def cvt(gt_path, save_path, img_folder):
"""
将icdar2015格式的gt转换为json格式
:param gt_path:
:param save_path:
:return:
"""
gt_dict = {'data_root': img_folder}
data_list = []
for file_path in tqdm(get_file_list(gt_path, p_postfix=['.txt'])):
content = load(file_path)
file_path = pathlib.Path(file_path)
img_name = file_path.name.replace('.txt', '.jpg')
cur_gt = {'img_name': img_name, 'annotations': []}
for line in content:
cur_line_gt = {'polygon': [], 'text': '', 'illegibility': False, 'language': 'Latin'}
chars_gt = [{'polygon': [], 'char': '', 'illegibility': False, 'language': 'Latin'}]
cur_line_gt['chars'] = chars_gt
line = line.split(',')
# 字符串级别的信息
x1, y1, x2, y2, x3, y3, x4, y4 = list(map(float, line[:8]))
cur_line_gt['polygon'] = [[x1, y1], [x2, y2], [x3, y3], [x4, y4]]
cur_line_gt['text'] = line[-1]
cur_line_gt['illegibility'] = True if cur_line_gt['text'] == '*' or cur_line_gt['text'] == '###' else False
cur_gt['annotations'].append(cur_line_gt)
data_list.append(cur_gt)
gt_dict['data_list'] = data_list
save(gt_dict, save_path)
def cvt(gt_path, save_path, imgs_folder):
gt_dict = {'data_root': imgs_folder}
data_list = []
ct = load(gt_path)
for img_id, anns in tqdm(ct.items()):
img_name = img_id.replace('gt', 'img') + '.jpg'
cur_gt = {'img_name': img_name, 'annotations': []}
for ann in anns:
cur_line_gt = {
'polygon': [],
'text': '',
'illegibility': False,
'language': 'Latin'
}
chars_gt = [{
'polygon': [],
'char': '',
'illegibility': False,
'language': 'Latin'
}]
cur_line_gt['chars'] = chars_gt
cur_line_gt['polygon'] = ann['points']
cur_line_gt['illegibility'] = ann['illegibility']
cur_gt['annotations'].append(cur_line_gt)
if len(cur_gt['annotations']) > 0:
data_list.append(cur_gt)
gt_dict['data_list'] = data_list
save(gt_dict, save_path)
print(len(gt_dict), len(data_list))
def load_data(self, json_path):
"""
从json文件中读取出 文本行的坐标和gt，字符的坐标和gt
:param json_path:
:return:
"""
content = load(json_path)
d = []
for gt in content['data_list']:
img_path = os.path.join(content['data_root'], gt['img_name'])
polygons = []
texts = []
illegibility_list = []
language_list = []
for annotation in gt['annotations']:
if len(annotation['polygon']) == 0:
continue
polygons.append(annotation['polygon'])
texts.append(annotation['text'])
illegibility_list.append(annotation['illegibility'])
language_list.append(annotation['language'])
for char_annotation in annotation['chars']:
if len(char_annotation['polygon']) == 0 or len(char_annotation['char']) == 0:
continue
polygons.append(char_annotation['polygon'])
texts.append(char_annotation['char'])
illegibility_list.append(char_annotation['illegibility'])
language_list.append(char_annotation['language'])
d.append({'img_path': img_path, 'polygons': np.array(polygons), 'texts': texts,
'illegibility': illegibility_list,
'language': language_list})
return d
def cvt_rec(gt_path, save_path, img_folder):
origin_gt = load(gt_path)
file_list = []
for img_name, gt in tqdm(origin_gt.items()):
assert len(gt) == 1
gt = gt[0]
img_path = os.path.join(img_folder, img_name + '.jpg')
file_list.append(img_path + '\t' + gt['transcription'] + '\t' + gt['language'])
save(file_list, save_path)
def cvt(gt_path, save_path, img_folder):
content = load(gt_path)
file_list = []
for line in tqdm(content):
line = line.split('\t')
img_path = os.path.join(img_folder, line[-2])
if not os.path.exists(img_path):
print(img_path)
file_list.append(img_path + '\t' + line[-1] + '\t' + 'Chinese')
# img = Image.open(img_path)
# plt.title(line[-1])
# plt.imshow(img)
# plt.show()
save(file_list, save_path)
def cvt(gt_path, save_path, img_folder):
content = load(gt_path)
file_list = []
for i, line in tqdm(enumerate(content)):
try:
line = line.split('.jpg ')
img_path = os.path.join(img_folder, line[-2])
file_list.append(img_path + '.jpg' + '\t' + line[-1] + '\t' +
'Chinese')
# img = Image.open(img_path)
# plt.title(line[-1])
# plt.imshow(img)
# plt.show()
except:
a = 1
save(file_list, save_path)
def cvt(gt_path, save_path, img_folder):
content = load(gt_path)
file_list = []
for line in tqdm(content):
img_relative_path = line.split(' ')[0]
img_path = os.path.join(img_folder, img_relative_path)
img_path = pathlib.Path(img_path)
label = img_path.stem.split('_')[1]
if not img_path.exists():
print(img_path)
file_list.append(str(img_path) + '\t' + label + '\t' + 'English')
# img = Image.open(img_path)
# plt.title(label)
# plt.imshow(img)
# plt.show()
save(file_list, save_path)
def cvt(gt_path, save_path, img_folder):
"""
将icdar2015格式的gt转换为json格式
:param gt_path:
:param save_path:
:return:
"""
gt_dict = {'data_root': img_folder}
data_list = []
for file_path in tqdm(get_file_list(gt_path, p_postfix=['.json'])):
content = load(file_path)
file_path = pathlib.Path(file_path)
img_name = file_path.stem + '.jpg'
cur_gt = {'img_name': img_name, 'annotations': []}
char_polygon_list, char_illegibility_list, char_text_list = decode_chars(
content['chars'])
for line in content['lines']:
cur_line_gt = {
'polygon': [],
'text': '',
'illegibility': False,
'language': 'Latin'
}
chars_gt = [{
'polygon': [],
'char': '',
'illegibility': False,
'language': 'Latin'
}]
cur_line_gt['chars'] = chars_gt
# 字符串级别的信息
cur_line_gt['polygon'] = np.array(line['points']).reshape(
-1, 2).tolist()
cur_line_gt['text'] = line['transcription']
cur_line_gt[
'illegibility'] = True if line['ignore'] == 1 else False
str_len = len(line['transcription'])
# 字符信息
flag = False
for char_idx in range(len(char_polygon_list)):
for str_idx in range(1, str_len + 1):
if ''.join(
char_text_list[char_idx:char_idx +
str_idx]) == line['transcription']:
chars_gt = []
for j in range(char_idx, char_idx + str_idx):
chars_gt.append({
'polygon':
char_polygon_list[j],
'char':
char_text_list[j],
'illegibility':
char_illegibility_list[j],
'language':
'Latin'
})
cur_line_gt['chars'] = chars_gt
char_polygon_list = char_polygon_list[char_idx +
str_len:]
char_text_list = char_text_list[char_idx + str_len:]
char_illegibility_list = char_illegibility_list[
char_idx + str_len:]
flag = True
break
if flag:
break
cur_gt['annotations'].append(cur_line_gt)
data_list.append(cur_gt)
gt_dict['data_list'] = data_list
save(gt_dict, save_path)
def __init__(self, txt_path, transform=None, target_transform=None):
self.data_list = load(txt_path)
self.transform = transform
self.target_transform = target_transform
