-
Notifications
You must be signed in to change notification settings - Fork 10
/
Copy pathtest.py
177 lines (161 loc) · 7.79 KB
/
test.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
import argparse
import os
import random
from datetime import datetime
import cv2
import imageio
import numpy as np
import staintools
import torch
from skimage.metrics import structural_similarity, peak_signal_noise_ratio
from torch.utils.data import DataLoader
from tqdm import tqdm
from models import StainNet, ResnetGenerator
from utils import list_file_tree, SingleImage
def detect_image(opt, model):
dataset = SingleImage(opt.source_dir)
dataloader = DataLoader(dataset,
batch_size=10,
num_workers=4,
drop_last=False)
file_list = dataset.image_list
save_path = os.path.join(opt.save_root, os.path.split(opt.model_path)[1][:-4])
num = 0
for imgs in tqdm(dataloader):
with torch.no_grad():
imgs = imgs.cuda()
imgs = (imgs - 0.5) * 2
outputs = model(imgs)
outputs = outputs * 0.5 + 0.5
outputs = outputs.clamp(0, 1).detach().cpu().numpy()
for out in outputs:
file_path = file_list[num]
file_path = os.path.join(save_path, os.path.split(file_path)[1])
os.makedirs(os.path.split(file_path)[0], exist_ok=True)
imageio.imwrite(file_path[:-4] + ".png",
(out * 255).astype(np.uint8).transpose((1, 2, 0)))
num += 1
return save_path
def traditional_methods(opt):
image_source = list_file_tree(opt.source_dir, "png")
image_target = list_file_tree(opt.gt_dir, "png")
image_source.sort()
image_target.sort()
if opt.method == "reinhard":
normalizer = staintools.ReinhardColorNormalizer()
else:
normalizer = staintools.StainNormalizer(method=opt.method)
if opt.random_target:
num = random.randint(0, len(image_target) - 1)
save_path = os.path.join(opt.save_root, opt.method + "_random")
os.makedirs(save_path, exist_ok=True)
print("target choose:", image_target[num])
target = staintools.read_image(image_target[num])
normalizer.fit(target)
for source in tqdm(image_source):
img = staintools.read_image(source)
filename = os.path.split(source)[1]
try:
img_normalized = normalizer.transform(img)
imageio.imwrite(os.path.join(save_path, filename[:-4] + ".png"),
img_normalized)
except:
print("error in ", source)
else:
save_path = os.path.join(opt.save_root, opt.method + "_matched")
os.makedirs(save_path, exist_ok=True)
for source, target in tqdm(zip(image_source, image_target)):
img1 = staintools.read_image(source)
img2 = staintools.read_image(target)
filename = os.path.split(source)[1]
try:
normalizer.fit(img2)
img1_normalized = normalizer.transform(img1)
imageio.imwrite(os.path.join(save_path, filename[:-4] + ".png"),
img1_normalized)
except:
print("error in ", source)
return save_path
def test_result(opt):
print(opt.result_dir, opt.source_dir, opt.gt_dir)
reslut_files = list_file_tree(opt.result_dir, "png")
source_files = list_file_tree(opt.source_dir, "png")
target_files = list_file_tree(opt.gt_dir, "png")
reslut_files.sort()
source_files.sort()
target_files.sort()
all_metirc = []
for reslut, source, target in tqdm(zip(reslut_files, source_files, target_files)):
image0 = imageio.imread(reslut)
image1 = imageio.imread(source)
image2 = imageio.imread(target)
ssim = structural_similarity(image0, image2, win_size=11, gaussian_weights=True, multichannel=True,
K1=0.01,
K2=0.03,
sigma=1.5, data_range=255)
psnr = peak_signal_noise_ratio(image0, image2, data_range=255)
image0 = cv2.cvtColor(image0, cv2.COLOR_RGB2GRAY)
image1 = cv2.cvtColor(image1, cv2.COLOR_RGB2GRAY)
image0 = image0.astype(np.float)
image1 = image1.astype(np.float)
image0 = (image0 - image0.min()) / (image0.max() - image0.min()) * 255
image1 = (image1 - image1.min()) / (image1.max() - image1.min()) * 255
ssim_source = structural_similarity(image0, image1, win_size=11, gaussian_weights=True, K1=0.01,
K2=0.03,
sigma=1.5, data_range=255)
all_metirc.append([source, {"ssim": ssim, "psnr": psnr, "ssim_source": ssim_source}])
mean_ssim = sum([k[1]["ssim"] for k in all_metirc]) / len(reslut_files)
mean_psnr = sum([k[1]["psnr"] for k in all_metirc]) / len(reslut_files)
mean_ssim_source = sum([k[1]["ssim_source"] for k in all_metirc]) / len(reslut_files)
torch.save(all_metirc, os.path.join(opt.result_dir, "all_metirc.data"))
print("SSIM GT", mean_ssim,
"PSNR GT", mean_psnr,
"SSIM Source", mean_ssim_source)
return mean_ssim, mean_psnr, mean_ssim_source
def test_methods(opt):
opt.save_root = os.path.split(opt.source_dir)[0]
if opt.method == "StainNet":
model = StainNet(opt.input_nc, opt.output_nc, opt.n_layer, opt.channels)
model = model.cuda()
model.load_state_dict(torch.load(opt.model_path))
model.eval()
opt.result_dir = detect_image(opt, model)
elif opt.method == "StainGAN":
model = ResnetGenerator(opt.input_nc, opt.output_nc, ngf=64, norm_layer=torch.nn.InstanceNorm2d, n_blocks=9)
model = model.cuda()
model.load_state_dict(torch.load(opt.model_path))
model.eval()
opt.result_dir = detect_image(opt, model)
elif opt.method in ["reinhard", "macenko", "vahadane"]:
opt.result_dir = traditional_methods(opt)
else:
raise RuntimeError("Not implemented Error!")
print("result save to ", opt.result_dir)
if opt.test_ssim:
mean_ssim, mean_psnr, mean_ssim_source = test_result(opt)
fs = open(os.path.join(opt.result_dir, "result.txt"), "a+")
fs.write(
"{}, SSIM GT:{}, PSNR GT:{}, SSIM Source:{}\n".format(datetime.now(), mean_ssim, mean_psnr,
mean_ssim_source))
print("test result save to ", os.path.join(opt.result_dir, "result.txt"))
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument("--source_dir", type=str, required=True,
help="path to source images for test")
parser.add_argument("--gt_dir", type=str, required=True,
help="path to ground truth images for test")
parser.add_argument("--method", default="StainNet", type=str,
help="different methods for test must be one of "
"{ StainNet StainGAN reinhard macenko vahadane khan }")
parser.add_argument('--test_ssim', action="store_true", default=True,
help='whether calculate SSIM , default is False')
parser.add_argument('--random_target', action="store_true", default=False,
help='random choose target or using matched ground truth, True is random choose target')
parser.add_argument('--input_nc', type=int, default=3, help='# of input image channels')
parser.add_argument('--output_nc', type=int, default=3, help='# of output image channels')
parser.add_argument('--channels', type=int, default=32, help='# of channels in StainNet')
parser.add_argument('--n_layer', type=int, default=3, help='# of layers in StainNet')
parser.add_argument('--model_path', type=str, required=True,
help='models path to load')
args = parser.parse_args()
test_methods(opt=args)