en2cn/Utils/tencent_hw.py

# -*- coding: utf-8 -*-
import json
import sys
import base64
import pandas as pd
import numpy as np
# import cv2
import re
import requests
import my_config
from tencentcloud.common import credential
from concurrent.futures.thread import ThreadPoolExecutor

from tencentcloud.common.profile.client_profile import ClientProfile
from tencentcloud.common.profile.http_profile import HttpProfile
from tencentcloud.common.exception.tencent_cloud_sdk_exception import TencentCloudSDKException
from tencentcloud.ocr.v20181119 import ocr_client, models


TENCENT_SK = {"LQY": {"ID": "AKIDBPOs761NcaK3YxrxveEXaOIkyxhCqdT2", "KEY": "mIabM1QhhiU1IIJ6dtEWdoKaJM8C1f5D"},
              "ZWJ": {"ID": "AKIDc2C9qIBpseBRSrrv06xceT2KMBl96Mcc", "KEY": "rTpyrUtnRoBHAnESRAhx2tTnumOlrgKh"},
              "WHQ": {"ID": "AKIDd2VieJvsnTqgLkCGcDajEUVF8mJZr0CC", "KEY": "w3vkJbKmytue9NKuDrGPK0kHXWEfopZW"},
              "ZY": {"ID": "AKIDNUMzLZcumOOAftdb5JIIFwMYVBbmooTr", "KEY": "FF4exDDfrmRpwmTxqSC2cvxioiNozuY8"},
              "GYW": {"ID": "AKIDHRJtI87z2lMhmPq8j2h4AyTgU3KEIK4W", "KEY": "2nHTO04uSecJHyNVNxFEbgUfogroUt6F"},
              "LWW": {"ID": "AKIDPYZy5GfImzEIY7q4QehGaeT5Q2PVv6Fk", "KEY": "MRIsBJA3g15PRDivra0DNWOzJN9ex60d"},
              "HJ": {"ID": "AKIDnKdhUVyn27ewzouOGsEuLXS7qVt2bTKV",  "KEY": "4P3SxCm7E4eyCBw85mFThgQuBwb0A1cy"}}


def tencent_handwriting(img):
    """
    每秒只能发<=10张
    :param img:
    :return:
    """
    cred = credential.Credential(TENCENT_SK["HJ"]["ID"],  # 改成自己的SecretId
                                 TENCENT_SK["HJ"]["KEY"])  # SecretKey
    httpProfile = HttpProfile()
    httpProfile.endpoint = "ocr.tencentcloudapi.com"
    clientProfile = ClientProfile(signMethod="TC3-HMAC-SHA256")  # 默认签名方法：HmacSHA256
    clientProfile.httpProfile = httpProfile
    client = ocr_client.OcrClient(cred, "ap-shanghai", clientProfile)
    imgbase64 = cimg(img)
    # print("imgbase64:", imgbase64)
    try:
        req = models.GeneralHandwritingOCRRequest()
        req.ImageBase64 = imgbase64
        print("图片占内存大小:【{}】".format(sys.getsizeof(req.ImageBase64) / (1024 ** 2)))
        resp = client.GeneralHandwritingOCR(req)
        res_str = resp.to_json_string()
        # print('结果:', res_str)
        # mongo_id = my_config.mycol.insert_one({'b64': imgbase64, 'result': res_str})
        try:
            res_dict = json.loads(res_str)
        except:
            res_dict = eval(res_str)
        textres = "".join(pd.DataFrame(res_dict["TextDetections"])["DetectedText"])
        print('结果:', textres)
        return textres, res_str
    except TencentCloudSDKException as err:
        print('ocr异常:', err)
        # mongo_id = my_config.mycol.insert_one({'b64': imgbase64, 'result': 'ocr异常:' + str(err)})
        return "", 'ocr异常:' + str(err)


def cimg(image):
    """
    将各来源的图片转为base64编码，先默认为本地图片路径
    :param image:
    :return:
    """
    if re.search("^https?:", str(image)):  # 线上图片
        filebyte = requests.get(image).content
    elif re.search("^[A-H]:", str(image)):   # 本地图片
        local_img = image
        filebyte = open(local_img, 'rb').read()
    else:  # 二进制文件流
        filebyte = image.read()
    # encoded = base64.b64encode(open(local_img, 'rb').read())

    base64_data = base64.b64encode(filebyte).rstrip().decode('utf-8')
    return base64_data


def tencent_ocr(imgbase64):
    # 通用手写体
    if type(imgbase64) is np.ndarray:  # 传进来的是np
        imgbase64 = np2base64(imgbase64)
    try:
        # print(bas64)
        cred = credential.Credential("AKIDnKdhUVyn27ewzouOGsEuLXS7qVt2bTKV", "4P3SxCm7E4eyCBw85mFThgQuBwb0A1cy")

        httpProfile = HttpProfile()
        httpProfile.endpoint = "ocr.tencentcloudapi.com"

        clientProfile = ClientProfile(signMethod="TC3-HMAC-SHA256")
        clientProfile.httpProfile = httpProfile
        client = ocr_client.OcrClient(cred, "ap-shanghai", clientProfile)

        # req = models.TableOCRRequest()
        req = models.GeneralHandwritingOCRRequest()
        req.ImageBase64 = imgbase64

        # resp = client.TableOCR(req)
        resp = client.GeneralHandwritingOCR(req)
        print('结果', json.loads(resp.to_json_string()))
        # mycol.insert_one({'b64': bas64, 'result': resp.to_json_string()})
        return resp.to_json_string()
    except TencentCloudSDKException as err:
        print('再次ocr异常:', err)
        return None


def np2base64(img_arr):
    """

    :param img_arr:
    :return:
    """
    # retval, buffer = cv2.imencode('.jpg', img_arr)
    # base64_str = base64.b64encode(buffer)
    # base64_str = base64_str.decode()
    # return base64_str


if __name__ == '__main__':
    aimg = r"G:\zwj\WL\en2cn\files\imgs\8-17.png"
    # aimg2 = "http://zxhx-pro-1302712961.cos.ap-beijing.myqcloud.com/aizj/d6d1a0b4ccd67643a1b8c51993b1b344_answer.png"
    res = tencent_handwriting(aimg)
    # imgbase64 = cimg(aimg2)
    # res = tencent_ocr(imgbase64)
    print("最后的结果：", res)

    import requests

    # res = requests.post("http://192.168.1.65:10116/E2CC",
    #                     files={"answer_hw": open(r"G:\zwj\WL\en2cn\files\7-18\3.jpg", "rb")})
    # f2 = open(r"G:\zwj\WL\en2cn\files\7-18\3.jpg", "rb")
    # print(type(cimg(f2.read())))