def command():
# Training settings
parser = argparse.ArgumentParser(description='PyTorch MNIST Example')
parser.add_argument('--batch-size',
type=int,
default=64,
metavar='N',
help='input batch size for training (default: 64)')
parser.add_argument('--test-batch-size',
type=int,
default=1000,
metavar='N',
help='input batch size for testing (default: 1000)')
parser.add_argument('--epochs',
type=int,
default=10,
metavar='N',
help='number of epochs to train (default: 10)')
parser.add_argument('--lr',
type=float,
default=0.01,
metavar='LR',
help='learning rate (default: 0.01)')
parser.add_argument('--momentum',
type=float,
default=0.5,
metavar='M',
help='SGD momentum (default: 0.5)')
parser.add_argument('--no-cuda',
action='store_true',
default=False,
help='disables CUDA training')
parser.add_argument('--seed',
type=int,
default=1,
metavar='S',
help='random seed (default: 1)')
parser.add_argument(
'--log-interval',
type=int,
default=10,
metavar='N',
help='how many batches to wait before logging training status')
args = parser.parse_args()
print(argsPrint(args))
return args
height = np.min([i.shape[0] for i in imgs])
end_pos = start_pos + img_width
if (ch == 1):
imgs = [i[:height, start_pos:end_pos] for i in imgs]
else:
imgs = [i[:height, start_pos:end_pos, :] for i in imgs]
imgs = [IMG.resize(i, rate) for i in imgs]
header_size = (30, int(img_width * rate), 3)
imgs = [titleInsert(i, t, header_size) for i, t in zip(imgs, text)]
return stackImages(imgs, thick=1, color=(0, 0, 0))
def main(args):
ch = IMG.getCh(args.channel)
imgs = [cv2.imread(name, ch) for name in args.image]
#text = ['[hitotsume]', '[futatsume]', '[mittsume]']
img = concat3Images(imgs, args.offset, args.img_width, args.channel,
args.img_rate)
cv2.imshow('test', img)
cv2.waitKey()
cv2.imwrite(getFilePath(args.out_path, 'concat', '.jpg'), img)
if __name__ == '__main__':
args = command()
argsPrint(args)
main(args)
def main(args):
# 画像を読み込む
imgs = [cv2.imread(name) for name in args.jpeg if IMG.isImgPath(name)]
# concatするためにすべての画像の高さを統一する
h = np.max([img.shape[0] for img in imgs])
imgs = [IMG.resize(img, h / img.shape[0]) for img in imgs]
# concatするためにすべての画像の幅を統一する
flg = cv2.BORDER_REFLECT_101
w = np.max([img.shape[1] for img in imgs])
imgs = [makeBorder(img, 0, 0, 0, w - img.shape[1], flg) for img in imgs]
# 画像に黒縁を追加する
flg = cv2.BORDER_CONSTANT
lw = args.line_width
imgs = [makeBorder(img, 0, lw, 0, lw, flg, (0, 0, 0)) for img in imgs]
# 縦横に連結するための画像リストと縦横情報を取得する
imgs, size = stackImgAndShape(imgs, args.row)
# 画像を連結してリサイズする
buf = [np.vstack(imgs[s]) for s in size]
img = IMG.resize(np.hstack(buf), args.resize)
# 連結された画像を保存する
name = F.getFilePath(args.out_path, 'concat', '.jpg')
print('save:', name)
cv2.imwrite(name, img)
if __name__ == '__main__':
args = command()
F.argsPrint(args)
main(args)