数学建模社区-数学中国

标题: Python-百度ocr识别 [打印本页]

作者: 檀俾九 时间: 2021-1-12 09:38
标题: Python-百度ocr识别

[code=python]#!/usr/bin/python
# -*- coding: utf-8 -*-

import base64
from os.path import exists
from tkinter import Tk
from tkinter.filedialog import askopenfilename
from urllib.parse import urlencode

import requests

def Dialog_box():
root = Tk()
root.withdraw()

filetypes = [(
      '图片文件(*.jpg;*.jpeg;*.gif;*.png;*.bmp)', '*.jpg;*.jpeg;*.gif;*.png;*.bmp'),
      ('所有文件(*.*)', '*.*')]

file_path = askopenfilename(filetypes=filetypes)
return file_path

def Read_file(file_path):
with open(file_path, mode='rb') as file_object:
      contents = file_object.read()
      return contents.rstrip()

def Baidu_ocr(binary):
url = "https://aip.baidubce.com/oauth/2.0/token?"
data = {
      "grant_type": "client_credentials",
      "client_id": "",  # 必填
      "client_secret": ""  # 必填
}
url += urlencode(data)
response = requests.get(url=url).json()
access_token = response['access_token']

url = "https://aip.baidubce.com/rest/2.0/ocr/v1/general_basic?access_token=" + access_token
post_data = {
      "image": base64.b64encode(binary),
      "url": "",
      "language_type": "CHN_ENG",
      "detect_direction": "false",
      "detect_language": "false",
      "probability": "false"
}
response = requests.post(url=url, data=post_data).json()
num = response['words_result_num']
result = ''
for i in range(num):
      result += response['words_result'][ i]['words'] + "\r\n"
return result

if __name__ == "__main__":
file_path = Dialog_box()
if exists(file_path):
      data = Read_file(file_path)
      print(Baidu_ocr(data))

[/code]

转发自派生社区

Python交流群：1047602540

欢迎光临数学建模社区-数学中国 (http://www.madio.net/)