python,将labelimg的标注文件json转换为paddleocr格式label.txt

可以使用以下代码将labelimg的标注文件json转换为paddleocr格式的label.txt文件：

import json
import os

def labelimg2paddleocr(json_path, label_path):
    with open(json_path, 'r', encoding='utf-8') as f:
        data = json.load(f)

    with open(label_path, 'w', encoding='utf-8') as f:
        for item in data:
            img_path = os.path.splitext(item['imagePath'])[0] + '.jpg'
            for shape in item['shapes']:
                label = shape['label']
                xmin = min(shape['points'], key=lambda x: x[0])[0]
                ymin = min(shape['points'], key=lambda x: x[1])[1]
                xmax = max(shape['points'], key=lambda x: x[0])[0]
                ymax = max(shape['points'], key=lambda x: x[1])[1]
                f.write(f'{img_path}\t{xmin},{ymin},{xmax},{ymax},{label}\n')

if __name__ == '__main__':
    json_path = 'path/to/labelimg/json/file'
    label_path = 'path/to/paddleocr/label.txt'
    labelimg2paddleocr(json_path, label_path)

其中，`json_path`为labelimg标注文件的json文件路径，`label_path`为转换后的paddleocr格式的label.txt文件路径。转换后的label.txt文件中每一行保存了对应图片中的一个目标的位置和标签信息。每个标注信息由5个字段组成，分别为图片路径、目标左上角x坐标、目标左上角y坐标、目标右下角x坐标、目标右下角y坐标和目标标签。