-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathto_coco.py
executable file
·286 lines (233 loc) · 9.93 KB
/
to_coco.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
'''
coco instance file example:
{
"info": {
"description": "COCO 2017 Dataset", # 数据集描述
"url": "http://cocodataset.org", # 下载地址
"version": "1.0", # 版本
"year": 2017, # 年份
"contributor": "COCO Consortium", # 提供者
"date_created": "2017/09/01" # 数据创建日期
},
"licenses": [
{
"url": "http://creativecommons.org/licenses/by-nc-sa/2.0/",
"id": 1,
"name": "Attribution-NonCommercial-ShareAlike License"
},
],
"images": [
{
"license": 4,
"file_name": "000000397133.jpg", # 图片名
"coco_url": "http://images.cocodataset.org/val2017/000000397133.jpg",# 网路地址路径
"height": 427, # 高
"width": 640, # 宽
"date_captured": "2013-11-14 17:02:52", # 数据获取日期
"flickr_url": "http://farm7.staticflickr.com/6116/6255196340_da26cf2c9e_z.jpg",# flickr网路地址
"id": 397133 # 图片的ID编号（每张图片ID是唯一的）
},
],
"categories": [ # 类别描述
{
"supercategory": "person", # 主类别
"id": 1, # 类对应的id （0 默认为背景）
"name": "person" # 子类别
},
{
"supercategory": "vehicle",
"id": 2,
"name": "bicycle"
},
{
"supercategory": "vehicle",
"id": 3,
"name": "car"
},
],
"annotations": [
{
"segmentation": [ # 对象的边界点（边界多边形）
[
224.24,297.18,# 第一个点 x,y坐标
228.29,297.18, # 第二个点 x,y坐标
234.91,298.29,
……
……
225.34,297.55
]
],
"area": 1481.3806499999994, # 区域面积
"iscrowd": 0, #
"image_id": 397133, # 对应的图片ID（与images中的ID对应）
"bbox": [217.62,240.54,38.99,57.75], # 定位边框 [x,y,w,h]
"category_id": 44, # 类别ID（与categories中的ID对应）
"id": 82445 # 对象ID，因为每一个图像有不止一个对象，所以要对每一个对象编号（每个对象的ID是唯一的）
},
]
}
'''
import sys
import os
import json
import cv2
import random
import shutil
class COCOCreater:
def __init__(self, src_dir, dst_dir):
self.train_map = {
"info": {
"description": "COCO 2017 Dataset", # 数据集描述
"url": "http://cocodataset.org", # 下载地址
"version": "1.0", # 版本
"year": 2017, # 年份
"contributor": "COCO Consortium", # 提供者
"date_created": "2017/09/01" # 数据创建日期
},
"licenses": [
{
"url": "http://creativecommons.org/licenses/by-nc-sa/2.0/",
"id": 1,
"name": "Attribution-NonCommercial-ShareAlike License"
},
],
"images": [
],
"categories": [ # 类别描述
],
"annotations": [
]
}
self.val_map = {
"info": {
"description": "COCO 2017 Dataset", # 数据集描述
"url": "http://cocodataset.org", # 下载地址
"version": "1.0", # 版本
"year": 2017, # 年份
"contributor": "COCO Consortium", # 提供者
"date_created": "2017/09/01" # 数据创建日期
},
"licenses": [
{
"url": "http://creativecommons.org/licenses/by-nc-sa/2.0/",
"id": 1,
"name": "Attribution-NonCommercial-ShareAlike License"
},
],
"images": [
],
"categories": [ # 类别描述
],
"annotations": [
]
}
self.support_formats=['jpg', 'JPG', 'png', 'PNG']
self.src_dir = src_dir
self.dst_dir = dst_dir
self.src_label_file = self.src_dir + '/label.txt'
self._create_dst_struct()
def read_ori_labels(self):
print("trans to coco: start read ori labels")
labels = []
with open(self.src_label_file, 'r') as f:
for line in f.readlines():
labels.append(line.strip('\r\n'))
random.shuffle(labels)
# self.ori_train_labels = labels[0: int(len(labels) * 0.8)]
# self.ori_val_labels = labels[int(len(labels) * 0.8):]
self.ori_train_labels = labels[0: ]
self.ori_val_labels = []
#print(self.ori_train_labels)
#print(self.ori_val_labels)
def create_train_map(self):
print("trans to coco: start create train dataset for coco")
self._create_coco_map(self.ori_train_labels, self.train_map, self.dst_dir_train2017)
with open(self.instances_train2017, "w") as f:
json.dump(self.train_map, f)
def create_val_map(self):
print("trans to coco: start create val data set for coco")
self._create_coco_map(self.ori_val_labels, self.val_map, self.dst_dir_val2017)
with open(self.instances_val2017, "w") as f:
json.dump(self.val_map, f)
def _create_coco_map(self, ori_labels, coco_map, img_dst_dir):
self.max_cls = -1
self.img_id = 0
self.box_id = 0
for i, line in enumerate(ori_labels):
print('trans to coco: %d/%d'%(i+1, len(ori_labels)))
self._create_by_line(line.strip('\r\n'), coco_map, img_dst_dir)
print('trans to coco: success')
#print(self.train_map)
#print(json.dumps(self.train_map))
def _create_by_line(self, line, coco_map, img_dst_dir):
fileds = line.split(' ')
img_name=fileds[0]
assert '.' in img_name, 'img_name do not has . '
assert img_name.split('.')[-1] in self.support_formats, 'img_name format is not illagle: (%s)'%img_name
img_path = os.path.join(self.src_dir, img_name)
img = cv2.imread(img_path)
assert img is not None, 'img is none, img path is:%s'%img_path
h, w, c = img.shape
shutil.copy(img_path, img_dst_dir)
#add image
self._add_image(img_name, h, w, coco_map)
#img box
if(len(fileds) > 4):
boxes = fileds[1:]
assert(len(boxes) % 5 == 0)
box_count = int(len(boxes) / 5)
for i in range(box_count):
box = boxes[i*5:i*5+5]
x0 = float(box[0])
y0 = float(box[1])
x1 = float(box[2])
y1 = float(box[3])
cls = int(float(box[4]))
self._add_cls(cls, coco_map)
self._add_box(x0,y0,x1,y1,cls, coco_map)
self.box_id += 1
self.img_id += 1
def _add_cls(self, cls, coco_map):
if cls > self.max_cls:
self.max_cls = cls
coco_map["categories"].append({"supercategory": "type_%d"%cls,
"id": self.max_cls,
"name": "type_%d"%cls})
def _add_image(self, img_name, h, w, coco_map):
coco_map["images"].append({"license": 1,
"file_name": "%s"%img_name,
"coco_url": "http://images.cocodataset.org/val2017/000000397133.jpg",
"height": h,
"width": w,
"date_captured": "2013-11-14 17:02:52",
"flickr_url": "http://farm7.staticflickr.com/6116/6255196340_da26cf2c9e_z.jpg",
"id": self.img_id
})
def _add_box(self, x0,y0,x1,y1,cls, coco_map):
box_area = (x1-x0)*(y1-y0)
coco_map["annotations"].append({"segmentation": [[x0, y0, x0, y1, x1,y1, x1, y0, x0, y0]],
"area": box_area,
"iscrowd": 0,
"image_id": self.img_id,
"bbox": [ x0, y0, x1-x0, y1-y0],
"category_id": cls,
"id": self.box_id
})
def _create_dst_struct(self):
assert(os.path.exists(self.dst_dir))
self.dst_dir_train2017 = os.path.join(self.dst_dir, 'train2017')
self.dst_dir_val2017 = os.path.join(self.dst_dir, 'val2017')
self.dst_dir_annotations = os.path.join(self.dst_dir, 'annotations')
self.instances_train2017 = os.path.join(self.dst_dir_annotations, 'instances_train2017.json')
self.instances_val2017 = os.path.join(self.dst_dir_annotations, 'instances_val2017.json')
if not os.path.exists(self.dst_dir_train2017):
os.mkdir(self.dst_dir_train2017)
if not os.path.exists(self.dst_dir_val2017):
os.mkdir(self.dst_dir_val2017)
if not os.path.exists(self.dst_dir_annotations):
os.mkdir(self.dst_dir_annotations)
if __name__ == '__main__':
coco = COCOCreater('./third_bridge_0', '../ttt')
coco.read_ori_labels()
coco.create_train_map()
coco.create_val_map()