convert_cityscapes_autolabeling.py

import os
import glob
import argparse

import numpy as np
from PIL import Image
import cityscapes_labels as cl


def parse_args():
    parser = argparse.ArgumentParser(
        description='Convert autolabelled data according to PaddleSeg data format'
    )
    parser.add_argument(
        "--dataset_root", dest="dataset_root", default=None, type=str)
    return parser.parse_args()


def get_color_map_list(num_classes):
    """ Returns the color map for visualizing the segmentation mask,
        which can support arbitrary number of classes.
    Args:
        num_classes: Number of classes
    Returns:
        The color map
    """
    color_map = num_classes * [0, 0, 0]
    for i in range(0, num_classes):
        j = 0
        lab = i
        while lab:
            color_map[i * 3] |= (((lab >> 0) & 1) << (7 - j))
            color_map[i * 3 + 1] |= (((lab >> 1) & 1) << (7 - j))
            color_map[i * 3 + 2] |= (((lab >> 2) & 1) << (7 - j))
            j += 1
            lab >>= 3

    return color_map


def main(args):
    dataset_root = args.dataset_root
    label_dir = os.path.join(dataset_root, 'refinement_final', 'train_extra')
    if dataset_root is None or not os.path.isdir(
            dataset_root) or not os.path.isdir(label_dir):
        raise ValueError(
            "The dataset is not Found or the folder structure is nonconfoumance."
        )

    autolabeling_label_files = sorted(
        glob.glob(os.path.join(label_dir, '*', '*_leftImg8bit.png')))
    print('start converting...')
    color_map = get_color_map_list(255)
    count = 0
    for file in autolabeling_label_files:
        mask = np.array(Image.open(file))
        for k, v in cl.label2trainid.items():
            binary_mask = (mask == k)
            mask[binary_mask] = v
        new_mask = Image.fromarray(mask.astype(np.uint8)).convert('P')
        new_mask.putpalette(color_map)
        file_split = file.split(os.sep)
        path = os.path.join(dataset_root, 'convert_autolabelled',
                            file_split[-2])
        if not os.path.exists(path):
            os.makedirs(path)
        new_mask.save(os.path.join(path, file_split[-1]))
        count += 1
        if count % 10 == 0:
            print('processed {} images'.format(count))


if __name__ == '__main__':
    args = parse_args()
    main(args)