label_ops.py 9.6 KB
Newer Older
W
WenmuZhou 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#    http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
from __future__ import unicode_literals

import numpy as np
T
tink2123 已提交
21
import string
W
WenmuZhou 已提交
22 23 24 25 26 27 28 29 30 31 32 33 34


class ClsLabelEncode(object):
    def __init__(self, label_list, **kwargs):
        self.label_list = label_list

    def __call__(self, data):
        label = data['label']
        if label not in self.label_list:
            return None
        label = self.label_list.index(label)
        data['label'] = label
        return data
W
WenmuZhou 已提交
35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55


class DetLabelEncode(object):
    def __init__(self, **kwargs):
        pass

    def __call__(self, data):
        import json
        label = data['label']
        label = json.loads(label)
        nBox = len(label)
        boxes, txts, txt_tags = [], [], []
        for bno in range(0, nBox):
            box = label[bno]['points']
            txt = label[bno]['transcription']
            boxes.append(box)
            txts.append(txt)
            if txt in ['*', '###']:
                txt_tags.append(True)
            else:
                txt_tags.append(False)
M
MissPenguin 已提交
56
        boxes = self.expand_points_num(boxes)
W
WenmuZhou 已提交
57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74
        boxes = np.array(boxes, dtype=np.float32)
        txt_tags = np.array(txt_tags, dtype=np.bool)

        data['polys'] = boxes
        data['texts'] = txts
        data['ignore_tags'] = txt_tags
        return data

    def order_points_clockwise(self, pts):
        rect = np.zeros((4, 2), dtype="float32")
        s = pts.sum(axis=1)
        rect[0] = pts[np.argmin(s)]
        rect[2] = pts[np.argmax(s)]
        diff = np.diff(pts, axis=1)
        rect[1] = pts[np.argmin(diff)]
        rect[3] = pts[np.argmax(diff)]
        return rect

M
MissPenguin 已提交
75 76 77 78 79 80 81 82 83 84 85
    def expand_points_num(self, boxes):
        max_points_num = 0
        for box in boxes:
            if len(box) > max_points_num:
                max_points_num = len(box)
        ex_boxes = []
        for box in boxes:
            ex_box = box + [box[-1]] * (max_points_num - len(box))
            ex_boxes.append(ex_box)
        return ex_boxes

W
WenmuZhou 已提交
86 87 88 89 90 91 92 93 94

class BaseRecLabelEncode(object):
    """ Convert between text-label and text-index """

    def __init__(self,
                 max_text_length,
                 character_dict_path=None,
                 character_type='ch',
                 use_space_char=False):
M
MissPenguin 已提交
95
        support_character_type = [
T
tink2123 已提交
96 97
            'ch', 'en', 'EN_symbol', 'french', 'german', 'japan', 'korean',
            'EN', 'it', 'xi', 'pu', 'ru', 'ar', 'ta', 'ug', 'fa', 'ur', 'rs',
T
tink2123 已提交
98 99
            'oc', 'rsc', 'bg', 'uk', 'be', 'te', 'ka', 'chinese_cht', 'hi',
            'mr', 'ne'
M
MissPenguin 已提交
100
        ]
W
WenmuZhou 已提交
101
        assert character_type in support_character_type, "Only {} are supported now but get {}".format(
M
MissPenguin 已提交
102
            support_character_type, character_type)
W
WenmuZhou 已提交
103 104

        self.max_text_len = max_text_length
T
tink2123 已提交
105 106
        self.beg_str = "sos"
        self.end_str = "eos"
W
WenmuZhou 已提交
107 108 109
        if character_type == "en":
            self.character_str = "0123456789abcdefghijklmnopqrstuvwxyz"
            dict_character = list(self.character_str)
T
tink2123 已提交
110
        elif character_type == "EN_symbol":
T
tink2123 已提交
111 112 113 114
            # same with ASTER setting (use 94 char).
            self.character_str = string.printable[:-6]
            dict_character = list(self.character_str)
        elif character_type in support_character_type:
W
WenmuZhou 已提交
115
            self.character_str = ""
T
tink2123 已提交
116 117
            assert character_dict_path is not None, "character_dict_path should not be None when character_type is {}".format(
                character_type)
W
WenmuZhou 已提交
118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145
            with open(character_dict_path, "rb") as fin:
                lines = fin.readlines()
                for line in lines:
                    line = line.decode('utf-8').strip("\n").strip("\r\n")
                    self.character_str += line
            if use_space_char:
                self.character_str += " "
            dict_character = list(self.character_str)
        self.character_type = character_type
        dict_character = self.add_special_char(dict_character)
        self.dict = {}
        for i, char in enumerate(dict_character):
            self.dict[char] = i
        self.character = dict_character

    def add_special_char(self, dict_character):
        return dict_character

    def encode(self, text):
        """convert text-label into text-index.
        input:
            text: text labels of each image. [batch_size]

        output:
            text: concatenated text index for CTCLoss.
                    [sum(text_lengths)] = [text_index_0 + text_index_1 + ... + text_index_(n - 1)]
            length: length of each text. [batch_size]
        """
W
WenmuZhou 已提交
146
        if len(text) == 0 or len(text) > self.max_text_len:
W
WenmuZhou 已提交
147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203
            return None
        if self.character_type == "en":
            text = text.lower()
        text_list = []
        for char in text:
            if char not in self.dict:
                # logger = get_logger()
                # logger.warning('{} is not in dict'.format(char))
                continue
            text_list.append(self.dict[char])
        if len(text_list) == 0:
            return None
        return text_list


class CTCLabelEncode(BaseRecLabelEncode):
    """ Convert between text-label and text-index """

    def __init__(self,
                 max_text_length,
                 character_dict_path=None,
                 character_type='ch',
                 use_space_char=False,
                 **kwargs):
        super(CTCLabelEncode,
              self).__init__(max_text_length, character_dict_path,
                             character_type, use_space_char)

    def __call__(self, data):
        text = data['label']
        text = self.encode(text)
        if text is None:
            return None
        data['length'] = np.array(len(text))
        text = text + [0] * (self.max_text_len - len(text))
        data['label'] = np.array(text)
        return data

    def add_special_char(self, dict_character):
        dict_character = ['blank'] + dict_character
        return dict_character


class AttnLabelEncode(BaseRecLabelEncode):
    """ Convert between text-label and text-index """

    def __init__(self,
                 max_text_length,
                 character_dict_path=None,
                 character_type='ch',
                 use_space_char=False,
                 **kwargs):
        super(AttnLabelEncode,
              self).__init__(max_text_length, character_dict_path,
                             character_type, use_space_char)

    def add_special_char(self, dict_character):
L
LDOUBLEV 已提交
204 205 206
        self.beg_str = "sos"
        self.end_str = "eos"
        dict_character = [self.beg_str] + dict_character + [self.end_str]
W
WenmuZhou 已提交
207 208
        return dict_character

L
LDOUBLEV 已提交
209 210
    def __call__(self, data):
        text = data['label']
W
WenmuZhou 已提交
211
        text = self.encode(text)
L
LDOUBLEV 已提交
212 213
        if text is None:
            return None
L
LDOUBLEV 已提交
214
        if len(text) >= self.max_text_len:
L
LDOUBLEV 已提交
215 216 217
            return None
        data['length'] = np.array(len(text))
        text = [0] + text + [len(self.character) - 1] + [0] * (self.max_text_len
T
tink2123 已提交
218
                                                               - len(text) - 2)
L
LDOUBLEV 已提交
219 220 221 222 223 224 225
        data['label'] = np.array(text)
        return data

    def get_ignored_tokens(self):
        beg_idx = self.get_beg_end_flag_idx("beg")
        end_idx = self.get_beg_end_flag_idx("end")
        return [beg_idx, end_idx]
W
WenmuZhou 已提交
226 227 228 229 230 231 232 233 234 235

    def get_beg_end_flag_idx(self, beg_or_end):
        if beg_or_end == "beg":
            idx = np.array(self.dict[self.beg_str])
        elif beg_or_end == "end":
            idx = np.array(self.dict[self.end_str])
        else:
            assert False, "Unsupport type %s in get_beg_end_flag_idx" \
                          % beg_or_end
        return idx
T
tink2123 已提交
236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257


class SRNLabelEncode(BaseRecLabelEncode):
    """ Convert between text-label and text-index """

    def __init__(self,
                 max_text_length=25,
                 character_dict_path=None,
                 character_type='en',
                 use_space_char=False,
                 **kwargs):
        super(SRNLabelEncode,
              self).__init__(max_text_length, character_dict_path,
                             character_type, use_space_char)

    def add_special_char(self, dict_character):
        dict_character = dict_character + [self.beg_str, self.end_str]
        return dict_character

    def __call__(self, data):
        text = data['label']
        text = self.encode(text)
T
tink2123 已提交
258
        char_num = len(self.character)
T
tink2123 已提交
259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281
        if text is None:
            return None
        if len(text) > self.max_text_len:
            return None
        data['length'] = np.array(len(text))
        text = text + [char_num] * (self.max_text_len - len(text))
        data['label'] = np.array(text)
        return data

    def get_ignored_tokens(self):
        beg_idx = self.get_beg_end_flag_idx("beg")
        end_idx = self.get_beg_end_flag_idx("end")
        return [beg_idx, end_idx]

    def get_beg_end_flag_idx(self, beg_or_end):
        if beg_or_end == "beg":
            idx = np.array(self.dict[self.beg_str])
        elif beg_or_end == "end":
            idx = np.array(self.dict[self.end_str])
        else:
            assert False, "Unsupport type %s in get_beg_end_flag_idx" \
                          % beg_or_end
        return idx