-
Notifications
You must be signed in to change notification settings - Fork 2
/
generate_density_map_perfect_names_SHAB_QNRF_NWPU_JHU.py
252 lines (220 loc) · 10.7 KB
/
generate_density_map_perfect_names_SHAB_QNRF_NWPU_JHU.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
import cv2
import numpy as np
import scipy
import scipy.io as scio
from PIL import Image
import time
import math
import os
import glob
def get_density_map_gaussian(H, W, ratio_h, ratio_w, points, adaptive_kernel=False, fixed_value=15):
h = H
w = W
density_map = np.zeros([h, w], dtype=np.float32)
num_gt = np.squeeze(points).shape[0]
if num_gt == 0:
return density_map
for idx, p in enumerate(points):
p = np.round(p).astype(int)
p[0], p[1] = min(h-1, math.floor(p[1] * ratio_h)), min(w-1, math.floor(p[0] * ratio_w))
sigma = fixed_value
sigma = max(1, sigma)
gaussian_radius = 7
gaussian_map = np.multiply(
cv2.getGaussianKernel(gaussian_radius*2+1, sigma),
cv2.getGaussianKernel(gaussian_radius*2+1, sigma).T
)
x_left, x_right, y_up, y_down = 0, gaussian_map.shape[1], 0, gaussian_map.shape[0]
# cut the gaussian kernel
if p[1] < 0 or p[0] < 0:
continue
if p[1] < gaussian_radius:
x_left = gaussian_radius - p[1]
if p[0] < gaussian_radius:
y_up = gaussian_radius - p[0]
if p[1] + gaussian_radius >= w:
x_right = gaussian_map.shape[1] - (gaussian_radius + p[1] - w) - 1
if p[0] + gaussian_radius >= h:
y_down = gaussian_map.shape[0] - (gaussian_radius + p[0] - h) - 1
density_map[
max(0, p[0]-gaussian_radius):min(h, p[0]+gaussian_radius+1),
max(0, p[1]-gaussian_radius):min(w, p[1]+gaussian_radius+1)
] += gaussian_map[y_up:y_down, x_left:x_right]
return density_map
def mkdir(path):
"""create a single empty directory if it didn't exist
Parameters:
path (str) -- a single directory path
"""
if not os.path.exists(path):
os.makedirs(path)
def mkdirs(paths):
"""create empty directories if they don't exist
Parameters:
paths (str list) -- a list of directory paths
"""
if isinstance(paths, list) and not isinstance(paths, str):
for path in paths:
mkdir(path)
else:
mkdir(paths)
# SHA: 300, 182
# SHB: 400, 316
if __name__ == "__main__":
is_train = 0 # 0 for test
# The processed data should be stored in folder with names silimar with `xx/train_data/`
train_test_for_gt_SH = 'train_data' if is_train else 'test_data'
train_test_for_gt_type2 = 'train' if is_train else 'test'
train_test_for_gt_type3 = 'train' if is_train else 'val'
train_test_for_den = 'den/train' if is_train else 'den/test'
dataset = 'QNRF_large'
if dataset == 'SHA':
# num_img = 300 if is_train else 182
image_dir_path = "ShanghaiTech/part_A_final/" + train_test_for_gt_SH + "/images"
ground_truth_dir_path = "ShanghaiTech/part_A_final/"+ train_test_for_gt_SH +"/ground_truth"
output_den_path = "./Processed_SHA_oriImg/" + train_test_for_den
# two extra paths
output_img_path = "./Processed_SHA_oriImg/ori/" + train_test_for_gt_SH + "/images"
output_mat_path = './Processed_SHA_oriImg/ori/' + train_test_for_gt_SH + "/ground_truth"
elif dataset == 'SHB':
# num_img = 400 if is_train else 316
image_dir_path = "ShanghaiTech/part_B_final/" + train_test_for_gt_SH + "/images"
ground_truth_dir_path = "ShanghaiTech/part_B_final/" + train_test_for_gt_SH + "/ground_truth"
output_den_path = "./Processed_SHB_oriImg/" + train_test_for_den
# two extra paths
output_img_path = "./Processed_SHB_oriImg/ori/" + train_test_for_gt_SH + "/images"
output_mat_path = './Processed_SHB_oriImg/ori/' + train_test_for_gt_SH + "/ground_truth"
elif dataset == 'QNRF_large':
# num_img = 1201 if is_train else 334
image_dir_path = "UCF-QNRF_ECCV18/" + train_test_for_gt_type2
ground_truth_dir_path = "UCF-QNRF_ECCV18/" + train_test_for_gt_type2
output_den_path = "./Processed_QNRF_large_oriImg/" + train_test_for_den
# two extra paths
output_img_path = "./Processed_QNRF_large_oriImg/ori/" + train_test_for_gt_SH + "/images" # using `gt_SH` here
output_mat_path = './Processed_QNRF_large_oriImg/ori/' + train_test_for_gt_SH + "/ground_truth"
# elif dataset == 'UCF50': # take all images as testing images
# num_img = 50
# image_dir_path = "UCF_CC_50/images/UCF_CC_50_img"
# ground_truth_dir_path = "UCF_CC_50/UCF_CC_50_mat"
# output_den_path = "./UCF50/" + train_test
elif dataset == 'NWPU_large':
image_dir_path = "NWPU/" + train_test_for_gt_type3 + "/imgs"
ground_truth_dir_path = "NWPU/" + train_test_for_gt_type3 + "/mats"
output_den_path = "./Processed_NWPU_large_oriImg/" + train_test_for_den
# two extra paths
output_img_path = "./Processed_NWPU_large_oriImg/ori/" + train_test_for_gt_SH + "/images"
output_mat_path = './Processed_NWPU_large_oriImg/ori/' + train_test_for_gt_SH + "/ground_truth"
elif dataset == 'JHU_large':
image_dir_path = "JHU_offical/" + train_test_for_gt_type2 + "/images"
ground_truth_dir_path = "JHU_offical/gt_npz/" + train_test_for_gt_type2
output_den_path = "./Processed_JHU_large_oriImg/" + train_test_for_den
# two extra paths
output_img_path = "./Processed_JHU_large_oriImg/ori/" + train_test_for_gt_SH + "/images"
output_mat_path = './Processed_JHU_large_oriImg/ori/' + train_test_for_gt_SH + "/ground_truth"
elif dataset == 'BDdata_large':
assert is_train == 1
image_dir_path = "BDdata/filter_imgs_100"
ground_truth_dir_path = "BDdata/filter_mats_100"
output_den_path = "./Processed_BDdata_large_oriImg/" + train_test_for_den
# two extra paths
output_img_path = "./Processed_BDdata_large_oriImg/ori/" + train_test_for_gt_SH + "/images"
output_mat_path = './Processed_BDdata_large_oriImg/ori/' + train_test_for_gt_SH + "/ground_truth"
else:
assert 1==2
mkdirs(output_den_path)
mkdirs(output_img_path)
mkdirs(output_mat_path)
img_paths = None
if dataset.find("QNRF") != -1: # as gt and images are in the same folder for QNRF dataset
img_paths = glob.glob(image_dir_path + "/*.jpg")
else:
img_paths = glob.glob(image_dir_path + "/*")
for img_path in img_paths:
if dataset == 'SHA' or dataset == 'SHB':
gt_path = os.path.join(ground_truth_dir_path, "GT_" + os.path.basename(img_path)[:-4] + ".mat")
elif dataset == 'QNRF' or dataset == 'QNRF_large':
gt_path = os.path.join(ground_truth_dir_path, os.path.basename(img_path)[:-4] + "_ann.mat")
# elif dataset == 'UCF50':
# img_path = os.path.join(image_dir_path, "img_"+("%d" % (i+1))+".jpg")
# gt_path = os.path.join(image_dir_path, "img_"+("%d" % (i+1))+"_ann.mat")
elif dataset == 'NWPU' or dataset == 'NWPU_large' or dataset == 'BDdata_large':
gt_path = os.path.join(ground_truth_dir_path, os.path.basename(img_path)[:-4] + ".mat")
elif dataset == 'JHU' or dataset == 'JHU_large':
gt_path = os.path.join(ground_truth_dir_path, os.path.basename(img_path)[:-4] + ".npz")
else:
assert 1==2
print('Processing img: ', img_path)
img = Image.open(img_path).convert('RGB')
height = img.size[1]
width = img.size[0]
if dataset == 'SHA' or dataset == 'SHB':
points = scio.loadmat(gt_path)['image_info'][0][0][0][0][0]
elif dataset.find('QNRF') != -1 or dataset == 'UCF50' or dataset.find('NWPU') != -1 or dataset.find('BDdata') != -1:
points = scio.loadmat(gt_path)['annPoints']
elif dataset.find('JHU') != -1:
points = np.load(gt_path)['loc_info']
else:
assert 1==2
resize_height = height
resize_width = width
if dataset == 'SHA' or dataset == 'UCF50':
if resize_height <= 416:
tmp = resize_height
resize_height = 416
resize_width = (resize_height / tmp) * resize_width
if resize_width <= 416:
tmp = resize_width
resize_width = 416
resize_height = (resize_width / tmp) * resize_height
resize_height = math.ceil(resize_height / 32) * 32
resize_width = math.ceil(resize_width / 32) * 32
elif dataset == 'SHB':
resize_height = 768
resize_width = 1024
elif dataset == 'QNRF':
resize_height = 768
resize_width = 1024
elif dataset == 'QNRF_large' or dataset == 'NWPU_large' or dataset == 'JHU_large' or dataset == 'BDdata_large':
if resize_width >= 2048:
tmp = resize_width
resize_width = 2048
resize_height = (resize_width / tmp) * resize_height
if resize_height >= 2048:
tmp = resize_height
resize_height = 2048
resize_width = (resize_height / tmp) * resize_width
if resize_height <= 416:
tmp = resize_height
resize_height = 416
resize_width = (resize_height / tmp) * resize_width
if resize_width <= 416:
tmp = resize_width
resize_width = 416
resize_height = (resize_width / tmp) * resize_height
# other constraints
if resize_height < resize_width:
if resize_width / resize_height > 2048/416:
resize_width = 2048
resize_height = 416
else:
if resize_height / resize_width > 2048/416:
resize_height = 2048
resize_width = 416
resize_height = math.ceil(resize_height / 32) * 32
resize_width = math.ceil(resize_width / 32) * 32
else:
assert 1==2
ratio_h = (resize_height) / (height)
ratio_w = (resize_width) / (width)
gt = get_density_map_gaussian(resize_height, resize_width, ratio_h, ratio_w, points, False, 4)
gt = np.reshape(gt, [resize_height, resize_width]) # transpose into w, h
# transfer gt to float16 to save storage
gt = gt.astype(np.float16)
# Three stuffs to store
# 1. images with new folders
os.system('cp '+ img_path + ' ./' + os.path.join(output_img_path, dataset + '_' + os.path.basename(img_path)))
# 2. save density maps
np.save(os.path.join(output_den_path, dataset + "_" + os.path.basename(img_path)[:-4] + ".npy"), gt) # some extensions are '.JPG', so...
# 3. save mats
scio.savemat(os.path.join(output_mat_path, dataset + "_" + os.path.basename(img_path)[:-4] + ".mat"), {'annPoints':points})
print("complete!")