merge_by_image_id.py

import urllib.request
import sys
import os
import json


def merge_by_image_id(import_file):
    merged_images = []
    f = open(import_file, 'r')
    json_data = json.load(f)
    images = json_data["images"]
    annotations = json_data["annotations"]
    for i in range(len(annotations)):
        image_id = annotations[i]["image_id"]
        label_id = annotations[i]["label_id"]
        urls = [x["url"][0] for x in images if x["image_id"] == image_id]
        url = urls[0] if len(urls) else ""
        image = {"image_id": image_id, "label_id": label_id, "url": url}
        print("image_id {0}".format(image_id))
        merged_images.append(image)
    merged_images_dict = {"images": merged_images}
    fw = open('merged_images.json', 'w')
    json.dump(merged_images_dict, fw, indent=2)


if __name__ == "__main__":
    import_file = sys.argv[1]
    merge_by_image_id(import_file)