hr-ocr-regbook/convert_json.py

from pathlib import Path

import requests
import json
import base64
from itertools import chain

url = 'http://192.168.199.249:38811'


def send_request(img_path, image_type=0):
    with open(img_path, 'rb') as f:
        img_str: str = base64.encodebytes(f.read()).decode('utf-8')
        data = {
            'image': img_str,
            'image_type': image_type
        }
        r = requests.post(f'{url}/ocr_system/regbook', json=data)
        print(r.json())
        return r.json()


def _parse_result(r):
    if r['status'] == '000':
        r = r['result']
        if r:
            del r['confidence']
        res = {}
        for k, v in r.items():
            if isinstance(v, dict):
                res[k] = v['text']
            else:
                res[k] = v
        return res
    elif r['status'] == '101':
        return "101"


if __name__ == '__main__':

    # 0
    # root = Path(__file__).parent
    # print(root)
    # img_paths = chain(
    #     *[Path(root / 'images/clean/0').rglob(f'*.{ext}') for ext in ['jpeg', 'jpg', 'png', 'JPG', 'PNG']])
    # for img_path in img_paths:
    #     print(img_path)
    #     r = send_request(img_path, 0)
    #     res = _parse_result(r)
    #     print(res)
    #     img_path: Path = img_path
    #     d = img_path.parent
    #     fn = img_path.stem + '.json'
    #
    #     with (d / fn).open('w', encoding='utf-8') as f:
    #         json.dump(res, f, ensure_ascii=False, indent=4)
    root = Path(__file__).parent
    print(root)
    img_paths = chain(
        *[Path(root / 'images/clean/1').rglob(f'*.{ext}') for ext in ['jpeg', 'jpg', 'png', 'JPG', 'PNG']])
    for img_path in img_paths:
        print(img_path)
        r = send_request(img_path, 1)
        res = _parse_result(r)
        print(res)
        img_path: Path = img_path
        d = img_path.parent
        fn = img_path.stem + '.json'

        with (d / fn).open('w', encoding='utf-8') as f:
            json.dump(res, f, ensure_ascii=False, indent=4)