def main():
parser = argparse.ArgumentParser()
parser.add_argument("--folder_with_files",
type=str,
default="../data/rozszerzone/v1/train/",
help="path to folder with images and labels")
parser.add_argument("--out_file",
type=str,
default="out.txt",
help="file with paths selected by the program")
parser.add_argument("--n_images",
type=int,
default=10,
help="quantity of images to select")
parser.add_argument("--class_id",
type=int,
default=3,
help="class id to search images for")
parser.add_argument("--n_classes",
type=int,
default=14,
help="quantity of classes in dataset")
parser.add_argument("--out_folder_with_selected_images",
type=str,
default="out",
help="file with paths selected by the program")
opt = parser.parse_args()
print(opt)
mypath = opt.folder_with_files
if mypath[-1] != "/":
mypath.append('/')
# all images
imageDataList = getValidImageDatas(mypath)
sack = Sack(imageDataList)
# filter images for only opt.class_id
#selected_type = sack.getOnlyLabeled(opt.class_id) # max 573 for stop sign
selected_type = sack.getLabeled(opt.class_id) # max 1215 for stop sign
# order images to the biggest images of element
selected_type.sort(reverse=True,
key=lambda imageData: imageData.getMaxSize())
print("n = ", len(selected_type))
topN = selected_type[0:opt.n_images]
s1 = Sack(topN)
s1.copyImagesToFolder(opt.out_folder_with_selected_images)
s1.saveToFile(opt.out_file, True)
