-
Notifications
You must be signed in to change notification settings - Fork 63
/
Copy pathData.py
209 lines (188 loc) · 9.44 KB
/
Data.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
import os
import xml.etree.ElementTree as ET
import cv2
import config as cfg
import codecs
import selectivesearch
import numpy as np
import math
import sys
import matplotlib
matplotlib.use('TkAgg')
import matplotlib.pyplot as plt
def resize_image(in_image, new_width, new_height, out_image=None, resize_mode=cv2.INTER_CUBIC):
img = cv2.resize(in_image, (new_width, new_height), resize_mode)
if out_image:
cv2.imwrite(out_image, img)
return img
def IOU(ver1, vertice2): # ver1:[xmin, ymin, w,h], verticel2:[xmin, ymin, xmax, ymax]
vertice1 = [ver1[0], ver1[1], ver1[0] + ver1[2], ver1[1] + ver1[3]]
lu = np.maximum(vertice1[0:2], vertice2[0:2])
rd = np.minimum(vertice1[2:], vertice2[2:])
intersection = np.maximum(0.0, rd - lu)
inter_square = intersection[0] * intersection[1]
square1 = (vertice1[2] - vertice1[0]) * (vertice1[3] - vertice1[1])
square2 = (vertice2[2] - vertice2[0]) * (vertice2[3] - vertice2[1])
union_square = np.maximum(square1 + square2 - inter_square, 1e-10)
return np.clip(inter_square / union_square, 0.0, 1.0)
def view_bar(message, num, total):
rate = num / total
rate_num = int(rate * 40)
rate_nums = math.ceil(rate * 100)
r = '\r%s:[%s%s]%d%%\t%d/%d' % (message, ">" * rate_num, " " * (40 - rate_num), rate_nums, num, total,)
sys.stdout.write(r)
sys.stdout.flush()
def show_rect(img_path, regions, test_index):
ind_to_class = dict(zip(range(1, len(cfg.Classes) + 1), cfg.Classes))
img = cv2.imread(img_path)
img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
for x, y, w, h, cls_ind, score in regions:
xmin, ymin, xmax, ymax = int(x), int(y), int(x + w), int(y + h)
if xmin <= 0: xmin = 1
if xmax >= cfg.Image_w: xmax = cfg.Image_w - 1
if ymin <= 0: ymin = 1
if ymax >= cfg.Image_h: ymax = cfg.Image_h - 1
message = ind_to_class[cls_ind]
rect = cv2.rectangle(
img, (xmin, ymin), (xmax, ymax), (0, 255, 0), 2)
font = cv2.FONT_HERSHEY_SIMPLEX
cv2.putText(img, message + ': ' + str(round(score, 2)), (xmin + 5, ymin - 20), font, 1, (255,255,255), 2)
if not os.path.exists(cfg.Test_output):
os.makedirs(cfg.Test_output)
out_path = os.path.join(cfg.Test_output, test_index+'test.jpg')
cv2.imwrite(out_path, img, [int(cv2.IMWRITE_PNG_COMPRESSION), 1])
plt.imshow(img)
plt.show()
class data(object):
def __init__(self, is_save=True):
self.all_list = cfg.All_list
self.train_list = cfg.Train_list
self.valid_list = cfg.Valid_list
self.annotation_path = cfg.Annotation_path
self.images_path = cfg.Images_path
self.processed_path = cfg.Processed_path
self.classes = cfg.Classes
self.class_num = cfg.Class_num
self.class_to_ind = dict(zip(self.classes, range(1, len(self.classes) + 1)))
self.image_w = cfg.Image_w
self.image_h = cfg.Image_h
self.is_save = is_save
self.batch_size = cfg.Batch_size
self.roi_threshold = cfg.Roi_threshold
self.train_images_index = []
self.cursor = 0
self.epoch = 0
if not os.path.exists(self.processed_path):
os.makedirs(self.processed_path)
if len(os.listdir(self.processed_path)) == 0:
self.generate_labels()
def generate_labels(self):
with codecs.open(self.all_list, 'r', 'utf-8') as f:
lines = f.readlines()
for num, image_idx in enumerate(lines):
ground_truth_dic = self.load_annotation(image_idx)
image_path = os.path.join(self.images_path, image_idx.strip() + '.jpg')
img = cv2.imread(image_path)
img_lbl, regions = selectivesearch.selective_search(img, scale=1000, sigma=0.9, min_size=1000)
labels = []
for r in regions:
x, y, w, h = r['rect']
proposal_vertice = [x + 1, y, x + w, y + h, w, h]
proposal_bbox = [x, y, (x + w - 1), (y + h - 1)]
label = np.zeros(self.class_num * 5 - 4, dtype=np.float32) # 假设包括背景有5类，0：5是判断类别，5：5+4*4=21 是位置框信息
iou_val=0
for ground_truth, class_idx in ground_truth_dic.items():
#ground_truth = list(ground_truth)
xmin=(2*ground_truth[0]-ground_truth[2])/2.0
ymin=(2*ground_truth[1]-ground_truth[3])/2.0
ground_truth=[xmin,ymin,ground_truth[2],ground_truth[3]]
iou_val = IOU(ground_truth, proposal_bbox)
px = float(proposal_vertice[0]) + float(proposal_vertice[4] / 2.0) # 中心点X
py = float(proposal_vertice[1]) + float(proposal_vertice[5] / 2.0) # 中心点Y
pw = float(proposal_vertice[4]) # w
ph = float(proposal_vertice[5]) # h
gx = float(ground_truth[0]) # 中心点X
gy = float(ground_truth[1]) # 中心点Y
gw = float(ground_truth[2]) # W
gh = float(ground_truth[3]) # H
if iou_val < self.roi_threshold :
label[0] = 1
elif iou_val > self.roi_threshold:
label[0] = 0
label[class_idx] = 1
label[self.class_num + (class_idx-1)*4 : self.class_num + (class_idx-1)*4 + 4] = \
[((gx - px) / pw), ((gy - py) / ph), (np.log(gw / pw)), (np.log(gh / ph))]
break
for i in range(len(proposal_bbox)):
proposal_bbox[i] = (proposal_bbox[i] / 16.0)
proposal_bbox.insert(0, 0)
proposal_bbox.insert(0, iou_val)
proposal_bbox.extend(label)
labels.append(proposal_bbox)
view_bar("Process image of %s" % image_path, num + 1, len(lines))
if self.is_save:
if not os.path.exists(self.processed_path): os.makedirs(self.processed_path)
np.save((os.path.join(self.processed_path, image_idx.split('.')[0].strip())
+ '_data.npy'), labels)
def load_annotation(self, image_idx):
iamge_annotion_path = os.path.join(self.annotation_path, image_idx.strip() + '.xml')
tree = ET.parse(iamge_annotion_path)
objs = tree.findall('object')
labels = {}
for obj in objs:
bbox = obj.find('bndbox')
x1 = float(bbox.find('xmin').text)
y1 = float(bbox.find('ymin').text)
x2 = float(bbox.find('xmax').text)
y2 = float(bbox.find('ymax').text)
class_name = obj.find('name').text.lower().strip()
if class_name != 'left':
cls_ind = self.class_to_ind[obj.find('name').text.lower().strip()]
boxes = [(x2 + x1) / 2.0, (y2 + y1) / 2.0, x2 - x1, y2 - y1]
labels[tuple(boxes)] = cls_ind
return labels
def get_batch(self):
if len(self.train_images_index) == 0:
print('load train list----------')
with open(self.train_list, 'r') as f:
for line in f.readlines():
image_index = line.strip()
self.train_images_index.append(image_index)
np.random.shuffle(self.train_images_index)
images = []
rois = []
label = []
for i in range(self.batch_size):
images_path = os.path.join(self.images_path, self.train_images_index[self.cursor] + '.jpg')
image = cv2.imread(images_path)
images.append(image)
labels = np.load(os.path.join(self.processed_path, self.train_images_index[self.cursor] + '_data.npy'))
labels = sorted(labels.tolist(), reverse=True)
select_num = min(cfg.Roi_num, len(labels))
for rois_label in labels[0:select_num]:
rois.append(
[rois_label[1] + i, int(rois_label[2])-1, int(rois_label[3])-1, int(rois_label[4])+1, int(rois_label[5])+1])
label.append((rois_label[6:]))
self.cursor += 1
if self.cursor >= len(self.train_images_index):
self.cursor = 0
self.epoch += 1
np.random.shuffle(self.train_images_index)
rois = np.array(rois)
label = np.array(label)
images = np.array(images)
return images, rois, label
def get_valid_batch(self, test_index):
valid_images_index = test_index
images_path = os.path.join(self.images_path, valid_images_index + '.jpg')
images = cv2.imread(images_path)
labels = np.load(os.path.join(self.processed_path, valid_images_index + '_data.npy'))
labels = sorted(labels.tolist(), reverse=True)
rois = []
label = []
for rois_label in labels:
rois.append([rois_label[1], (rois_label[2]), (rois_label[3]), (rois_label[4]), (rois_label[5])])
label.append((rois_label[6:]))
rois = np.array(rois)
label = np.array(label)
return [images], rois, label