import json

with open("datasets/coco_custom/annotations/instances_val2017.json", "r") as f:

data = json.load(f)

# images, annotations, categories

print(data["images"][0])

print(data["categories"][0])

print(data["annotations"][0])

# format for ultralytics: class x_center y_center width height

# xywh format(0-1)

# divide width and xcenter by img width

# divide height and ycenter by img height

images_r = {}

for img in data["images"]:

images_r[img["id"]] = {

"file_name": img["file_name"],

"width": img["width"],

"height": img["height"],

}

for annotation in data["annotations"]:

img_id = annotation["image_id"]

left, top, boxwidth, boxheight = annotation["bbox"]

category_id = annotation["category_id"]

if category_id > 79:

continue

xcenter = left + (boxwidth / 2)

ycenter = top + (boxheight / 2)

image_r = images_r[img_id]

xcenter /= image_r["width"]

boxwidth /= image_r["width"]

ycenter /= image_r["height"]

boxheight /= image_r["height"]

file_name: str = image_r["file_name"]

file_name = file_name.replace(".jpg", ".txt")

msg = f"{category_id} {xcenter} {ycenter} {boxwidth} {boxheight}\n"

FILE = f"datasets/coco_custom/labels/val/{file_name}"

try:

with open(FILE, "a") as f:

f.write(msg)

except FileNotFoundError:

with open(FILE, "w") as f:

f.write(msg)

# for category in data["categories"]:

# print(f"{category['id']}: {category['name']}")