python 腾讯API 名片识别 通用图片转文字

记录名片识别和图片转文字。
腾讯AI开放平台提供的名片OCR很强大， 而且可以免费使用, 所以本文使用的是腾讯AI开放平台。
腾讯AI开放平台API调用需要注意签名方法, 请查看之前文章传送门

名片OCR API

import logging
logger = logging.getLogger(__name__)
import base64
import time
import requests
import json
def bc_ocr(_path = None, _byte = None):
    ''' 腾讯AI名片识别
    _path和_byte必须存在一个, 优先级 _byte > _path
    :param _path: 
    :param _bytes: 
    :return: 
    '''
    try:
        if _path is None and _byte is None: return
        if _byte is None:
            with open(_path, 'rb') as f:
                _byte = f.read()
        if _byte is None: return
        APP_ID = '你的 App ID'
        API_KEY = '你的 Api Key'
        b_str = base64.b64encode(_byte)
        params = {'app_id': APP_ID, 'time_stamp':int(time.time()), 'nonce_str': 'fa577ce340859f9fe', 'image': b_str}
        sign = get_sign_code(params, API_KEY)
        params['sign'] = sign
        url = 'https://api.ai.qq.com/fcgi-bin/ocr/ocr_bcocr'
        rt = requests.post(url, data=params)
        jsonData = json.loads(rt.text)
        logger.info('腾讯名片OCR结果：[{}]'.format(jsonData))
        if jsonData.get('ret') == 0:
            return jsonData.get('data')
    except Exception as e:
        logger.error('tencen business_card_ocr error [{}]'.format(e))

通用图片识别 API

import logging
logger = logging.getLogger(__name__)
import base64
import time
import requests
import json
def ocr(_path = None, _byte = None):
    '''腾讯 图片识别
    _path和_byte必须存在一个, 优先级 _byte > _path
    :param _path: 
    :param _byte: 
    :return: 
    '''
    try:
        if _path is None and _byte is None: return
        if _byte is None:
            with open(_path, 'rb') as f:
                _byte = f.read()
        if _byte is None: return
        APP_ID = '你的 App ID'
        API_KEY = '你的 Api Key'
        image_str = base64.b64encode(_byte)
        url = 'https://api.ai.qq.com/fcgi-bin/ocr/ocr_generalocr'
        params = {'app_id': APP_ID, 'time_stamp':int(time.time()), 'nonce_str': 'fa577ce340859f9fe', 'image': image_str}
        sign = get_sign_code(params, API_KEY)
        params['sign'] = sign
        rt = requests.post(url, data=params)
        jsonData = json.loads(rt.text)
        if jsonData.get('ret') == 0:
            item_list = jsonData.get('data').get('item_list')
            return '\r\n'.join([item.get('itemstring') for item in item_list])
    except Exception as e:
        logger.error('tencen ocr error [{}]'.format(e))