optimizer = optim.SGD(net.parameters(), lr=0.001,
momentum=0.9) #select the optimizer
exp_lr_scheduler = lr_scheduler.StepLR(optimizer, step_size=50, gamma=0.1)
# create the train_dataset_loader and val_dataset_loader.
cloud_data = CloudDataset(img_dir='data/images/', labels_dir='data/GTmaps/')
trainer = Trainer('inference', optimizer, exp_lr_scheduler, net, cfig, './log')
trainer.load_weights(trainer.find_last())
#trainer.load_weights('log/renset20190102T1348/model_renset_0046.pt')
since = time.time()
for x in range(0, 801, 5):
images = cloud_data[x]['image']
gt_map = cloud_data[x]['gt_map']
mask = trainer.detect(images)
mask = np.round(mask * 255)
# images=cv2.cvtColor(images,cv2.COLOR_BGR2GRAY)
# cv2.imwrite('result/{}_image.png'.format(x),images)
# cv2.imwrite('result/{}gt_map.png'.format(x),gt_map)
#cv2.imwrite('result/{}sigmoid.png'.format(x),mask)
#fig.set_size_inches(600/100.0,600/100.0)#输出width*height像素
print(mask.shape)
fig = plt.figure()
fig.set_size_inches(600 / 100.0, 300 / 100.0) #输出width*height像素
plt.subplot(121)
plt.xticks([]) #去掉横坐标值
plt.yticks([]) #去掉纵坐标值
