python对接百度通用ocr

前言

仅用于个人python学习记录,百度有官方的模块

pip install baidu-aip

access_token获取

略

代码

import requests
import base64
import json
import logging
from enum import Enum, unique
import keyring

logging.basicConfig()
logging.getLogger().setLevel(logging.DEBUG)

@unique
class OcrType(Enum):
    #标准版
    STANDARD_BASIC = "https://aip.baidubce.com/rest/2.0/ocr/v1/general_basic"
    #标准版含位置
    STANDARD_WITH_LOCATION = "https://aip.baidubce.com/rest/2.0/ocr/v1/general"
    #高精度版
    ACCURATE_BASIC = "https://aip.baidubce.com/rest/2.0/ocr/v1/accurate_basic"
    #高精度版含位置
    ACCURATE_WITH_LOCATION = "https://aip.baidubce.com/rest/2.0/ocr/v1/accurate"
    #办公文档识别
    DOC_ANALYSiS_OFFICE = "https://aip.baidubce.com/rest/2.0/ocr/v1/doc_analysis_office"
    #网络图片文字识别
    WEB_IMAGE = "https://aip.baidubce.com/rest/2.0/ocr/v1/webimage"
    #网络图片文字识别 含位置
    WEB_IMAGE_WITH_LOCATION = "https://aip.baidubce.com/rest/2.0/ocr/v1/webimage_loc"
    #手写文字识别
    HAND_WRITING = "https://aip.baidubce.com/rest/2.0/ocr/v1/handwriting"
    #数字识别
    NUMBERS = "https://aip.baidubce.com/rest/2.0/ocr/v1/numbers"
    #表格文字识别(同步接口)
    FORM_SYNCH = "https://aip.baidubce.com/rest/2.0/ocr/v1/form"
    #表格文字识别(异步接口)
    FORM_ASYNCH = "https://aip.baidubce.com/rest/2.0/solution/v1/form_ocr/request"
    #二维码识别
    QRCORD = "https://aip.baidubce.com/rest/2.0/ocr/v1/qrcode"

class BaiduGeneralOcr():
    #ocr版本
    ocr_type = 0
    
    def __init__(self,ocr_type:OcrType):
        '''
        ocr_type ocr识别类型 STANDARD_BASIC标准版 标准版含位置STANDARD_WITH_LOCATION 高精读版ACCURATE_BASIC 高精度版含位置ACCURATE_WITH_LOCATION

        '''
        self.ocr_type = ocr_type.value

    def gen_access_token(self,API_Key,Secret_Key):
        '''
        生成access_token
        API_Key API_KEY
        Secret_Key Secret_Key
        '''
        requestUrl = f'''https://aip.baidubce.com/oauth/2.0/token?grant_type=client_credentials&client_id={API_Key}&client_secret={Secret_Key}'''
        response = requests.get(requestUrl)
        if response:
            self.access_token = response.json()['access_token']
        
    def __check(self):
        '''
        参数检查
        '''
        image=self.image
        url=self.url
        pdf_file=self.pdf_file

        if((image==None or image=="") and (url==None or url=="") and (pdf_file==None or pdf_file=="")):
            raise ValueError("image,url,pdf_file至少传入一项")
        if(image!=None and image!=""):
            f = open(image, 'rb')
            self.options['image'] = base64.b64encode(f.read()).decode()
        elif((image==None or image=="") and (url!=None and url!="")):
            self.options['url'] = url
        else:
            f = open(pdf_path, 'rb')
            self.options['pdf_file'] = base64.b64encode(f.read()).decode()

        keys = self.options.keys()
        if(("image" not in keys) and ("url" not in keys) and ("pdf_file" not in keys)):
            raise ValueError("image,url,pdf_file至少传入一项")
        try:
            self.ocr_type.value
        except Exception as e:
            self.ocr_type = OcrType.STANDARD_BASIC
            logging.info("ocr_type类型,已重置为标准版")
    
    def __request(self,request_url,data):
        #设置header
        headers = {'content-type': 'application/x-www-form-urlencoded'}
        #请求
        print(self.options)
        print(headers)
        return requests.post(request_url+"?access_token=%s" %self.access_token, data= self.options, headers=headers).json()

    def recoginze(self,image:str=None,url:str=None,pdf_file:str=None,options=None):
        '''
        识别
        '''
        self.image=image
        self.url=url
        self.pdf_file=pdf_file

        self.options = options or {}
        self.__check()
        #发送请求
        return self.__request(self.ocr_type.value,options)

    def set_access_token(self,access_token):
        '''
        设置access_token
        access_token access_token
        '''
        self.access_token = access_token

    def ocr_help(self):
        print("==========百度ocr使用说明==========")
        print("本API基于ocr通用识别api编写,官方文档地址:https://cloud.baidu.com/doc/OCR/s/zk3h7xz52")
        print("使用示例:")
        print('''
        baiduOcr = BaiduGeneralOcr(OcrType.STANDARD_BASIC)
        baiduOcr.set_access_token("access_token")
        wordsList = baiduOcr.recoginze(image="D:\\txt1.png",options={})['words_result']
        for word in wordsList:
            print(word)
        ''')

baiduOcr = BaiduGeneralOcr(OcrType.STANDARD_BASIC)
baiduOcr.set_access_token(keyring.get_password("baidu_dev","access_token"))
response = baiduOcr.recoginze(image="D:\\txt1.png",options={})
wordsList = response['words_result']
for word in wordsList:
    print(word)