-
Notifications
You must be signed in to change notification settings - Fork 10
/
create_tf_record.py
304 lines (261 loc) · 10.9 KB
/
create_tf_record.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
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
146
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
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
"""
Convert dataset to TFRecord for TF object detection training.
Example usage:
python3 create_tf_record.py \
--root_dir ./ \
--image_dir images \
--annotation_dir annotations \
--output_dir tf-record \
--dataset_name radar-ml
Only datset_name is required.
Based on:
https://github.com/tensorflow/models/blob/master/research/object_detection/dataset_tools/create_pascal_tf_record.py
Copyright (c) 2019~2020 Lindo St. Angel
"""
import hashlib
import io
import logging
import os
import random
import re
import contextlib2
import numpy as np
import PIL.Image
import tensorflow as tf
import argparse
from lxml import etree
from object_detection.dataset_tools import tf_record_creation_util
from object_detection.utils import dataset_util
from object_detection.utils import label_map_util
logger = logging.getLogger(__name__)
NUM_TFRECORD_SHARDS = 1
TRAIN_VAL_SPLIT = 0.8
TFRECORD_TRAIN_NAME = 'train'
TFRECORD_VAL_NAME = 'val'
ALT_NAME_MAP = {
'lindo': 'person',
'nikki': 'person',
'eva': 'person',
'nico': 'person',
'unknown': 'person',
'polly': 'dog',
'rebel': 'cat',
'jack': 'cat'
}
def dict_to_tf_example(data,
label_map_dict,
image_subdirectory,
use_alt_names=False,
ignore_difficult_instances=False):
"""Convert XML derived dict to tf.Example proto.
Notice that this function normalizes the bounding box coordinates provided
by the raw data.
Args:
data: dict holding PASCAL XML fields for a single image (obtained by
running dataset_util.recursive_parse_xml_to_dict)
label_map_dict: A map from string label names to integers ids.
image_subdirectory: String specifying subdirectory within the
Pascal dataset directory holding the actual image data.
ignore_difficult_instances: Whether to skip difficult instances in the
dataset (default: False).
use_alt_names: Use class names that may be different than labels in images.
A translation map must be provided (default: False).
Returns:
example: The converted tf.Example.
Raises:
ValueError: if the image pointed to by data['filename'] is not a valid JPEG
"""
img_path = os.path.join(image_subdirectory, data['filename'])
with tf.io.gfile.GFile(img_path, 'rb') as fid:
encoded_jpg = fid.read()
encoded_jpg_io = io.BytesIO(encoded_jpg)
image = PIL.Image.open(encoded_jpg_io)
if image.format != 'JPEG':
raise ValueError('Image format must be JPEG.')
key = hashlib.sha256(encoded_jpg).hexdigest()
width = int(data['size']['width'])
height = int(data['size']['height'])
xmins = []
ymins = []
xmaxs = []
ymaxs = []
classes = []
classes_text = []
truncated = []
poses = []
difficult_obj = []
if 'object' in data:
for obj in data['object']:
difficult = bool(int(obj['difficult']))
if ignore_difficult_instances and difficult:
continue
difficult_obj.append(int(difficult))
xmin = float(obj['bndbox']['xmin'])
xmax = float(obj['bndbox']['xmax'])
ymin = float(obj['bndbox']['ymin'])
ymax = float(obj['bndbox']['ymax'])
xmins.append(xmin / width)
ymins.append(ymin / height)
xmaxs.append(xmax / width)
ymaxs.append(ymax / height)
#class_name = get_class_name_from_filename(data['filename'])
if use_alt_names:
class_name = ALT_NAME_MAP.get(obj['name'], obj['name'])
else:
class_name = obj['name']
print(class_name, label_map_dict[class_name])
classes_text.append(class_name.encode('utf8'))
classes.append(label_map_dict[class_name])
truncated.append(int(obj['truncated']))
poses.append(obj['pose'].encode('utf8'))
feature_dict = {
'image/height': dataset_util.int64_feature(height),
'image/width': dataset_util.int64_feature(width),
'image/filename': dataset_util.bytes_feature(
data['filename'].encode('utf8')),
'image/source_id': dataset_util.bytes_feature(
data['filename'].encode('utf8')),
'image/key/sha256': dataset_util.bytes_feature(key.encode('utf8')),
'image/encoded': dataset_util.bytes_feature(encoded_jpg),
'image/format': dataset_util.bytes_feature('jpeg'.encode('utf8')),
'image/object/bbox/xmin': dataset_util.float_list_feature(xmins),
'image/object/bbox/xmax': dataset_util.float_list_feature(xmaxs),
'image/object/bbox/ymin': dataset_util.float_list_feature(ymins),
'image/object/bbox/ymax': dataset_util.float_list_feature(ymaxs),
'image/object/class/text': dataset_util.bytes_list_feature(classes_text),
'image/object/class/label': dataset_util.int64_list_feature(classes),
'image/object/difficult': dataset_util.int64_list_feature(difficult_obj),
'image/object/truncated': dataset_util.int64_list_feature(truncated),
'image/object/view': dataset_util.bytes_list_feature(poses),
}
return tf.train.Example(features=tf.train.Features(feature=feature_dict))
def create_tf_record(output_filename,
num_shards,
label_map_dict,
annotations_dir,
image_dir,
examples,
use_alt_names):
"""Creates a TFRecord file from examples.
Args:
output_filename: Path to where output file is saved.
num_shards: Number of shards for output file.
label_map_dict: The label map dictionary.
annotations_dir: Directory where annotation files are stored.
image_dir: Directory where image files are stored.
examples: Examples to parse and save to tf record.
use_alt_names: use alternative class name mapping.
"""
with contextlib2.ExitStack() as tf_record_close_stack:
output_tfrecords = tf_record_creation_util.open_sharded_output_tfrecords(
tf_record_close_stack, output_filename, num_shards)
for idx, example in enumerate(examples):
if idx % 10 == 0:
logger.info('On image %d of %d', idx, len(examples))
xml_path = os.path.join(annotations_dir, 'xmls', example + '.xml')
if not os.path.exists(xml_path):
logger.warning('Could not find %s, ignoring example.', xml_path)
continue
with tf.io.gfile.GFile(xml_path, 'r') as fid:
xml_str = fid.read()
xml = etree.fromstring(xml_str)
data = dataset_util.recursive_parse_xml_to_dict(xml)['annotation']
try:
tf_example = dict_to_tf_example(
data=data,
label_map_dict=label_map_dict,
image_subdirectory=image_dir,
use_alt_names=use_alt_names)
if tf_example:
shard_idx = idx % num_shards
output_tfrecords[shard_idx].write(tf_example.SerializeToString())
except ValueError:
logger.warning('Invalid example: %s, ignoring.', xml_path)
def gen_trainval_list(images_path):
"""Creates a list of image names without file extensions.
The list items will not match the ordering of the images on disk.
Args:
images_path: Path to where images are located.
"""
def make(file):
if file.endswith('.jpg' or '.jpeg'):
return os.path.basename(file).split('.')[0]
return [make(file) for file in os.listdir(images_path)]
def main(args):
logger.info('Reading dataset info.')
image_dir = os.path.join(args.root_dir, args.image_dir,
args.dataset_name)
logger.info(f'Image directory: {image_dir}')
annotations_dir = os.path.join(args.root_dir, args.annotation_dir,
args.dataset_name)
logger.info(f'Annotation directory: {annotations_dir}')
label_map = os.path.join(args.root_dir, args.annotation_dir,
args.dataset_name, args.label_map_name)
logger.info(f'Label map: {label_map}')
use_alt_names = args.use_alt_names
logger.info(f'use alt names: {use_alt_names}')
# Split data into training and validation sets.
random.seed(42)
examples_list = gen_trainval_list(image_dir)
random.shuffle(examples_list)
num_examples = len(examples_list)
num_train = int(TRAIN_VAL_SPLIT * num_examples)
train_examples = examples_list[:num_train]
val_examples = examples_list[num_train:]
logger.info('Found %d training and %d validation examples.',
len(train_examples), len(val_examples))
train_output_path = os.path.join(args.root_dir, args.output_dir,
args.dataset_name, TFRECORD_TRAIN_NAME)
val_output_path = os.path.join(args.root_dir, args.output_dir,
args.dataset_name, TFRECORD_VAL_NAME)
label_map_dict = label_map_util.get_label_map_dict(label_map)
# Create training TFRecord.
logger.info('Creating training TFRecord.')
create_tf_record(
train_output_path,
NUM_TFRECORD_SHARDS,
label_map_dict,
annotations_dir,
image_dir,
train_examples,
use_alt_names)
logger.info(f'Created training TFRecord: {train_output_path}')
# Create validation TFRecord.
logger.info('Creating validation TFRecord.')
create_tf_record(
val_output_path,
NUM_TFRECORD_SHARDS,
label_map_dict,
annotations_dir,
image_dir,
val_examples,
use_alt_names)
logger.info(f'Created validation TFRecord: {val_output_path}')
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--root_dir', type=str,
help='Root directory.',
default='./')
parser.add_argument('--output_dir', type=str,
help='TFRecord directory.',
default='tf-record')
parser.add_argument('--annotation_dir', type=str,
help='Annotation directory.',
default='annotations')
parser.add_argument('--label_map_name', type=str,
help='Label map name.',
default='label_map.pbtxt')
parser.add_argument('--image_dir', type=str,
help='Image directory.',
default='images')
parser.add_argument('--dataset_name', type=str,
help='Name of dataset',
required=True)
parser.add_argument('--use_alt_names', action='store_true',
help='Use alternative class names. Must match label_map_name.pbtxt')
parser.set_defaults(use_alt_names=False)
args = parser.parse_args()
logging.basicConfig(
format='%(asctime)s %(name)-12s %(levelname)-8s %(message)s',
level=logging.DEBUG)
main(args)