百度API文字识别配合selenium自动登录

识别图片验证码自动登录demo

1. 使用selenium需要下载chromedriver 网址: http://npm.taobao.org/mirrors/chromedriver/ 下载跟自己浏览器版本号最接近的版本.

2. https://cloud.baidu.com/ 百度开发平台. https://cloud.baidu.com/doc/OCR/s/zk3h7xz52 文字识别开发文档

3. 导包

   import random
   from selenium import webdriver
   import time
   import requests
   import json
   import base64
   import urllib
   

4. 建立浏览器对象和抓取验证码图片

   # 建立浏览器对象
   browser = webdriver.Chrome('D:\python\python6\chromedriver.exe')
   
   # 打开网址
   browser.get('http://localhost:8080/register')
   
   # 等待5秒
   time.sleep(3)
   
   # 选择元素
   myimg = browser.find_element_by_xpath('//*[@id="app"]/div/section/div[2]/table/tr[3]/td[2]/img')
   
   # 截取元素图
   myimg.screenshot('register.png')
   # 随机生成用户名与密码
   num = '0123456789qwertyuiopasdfghjklzxcvbnm'
   
   nums = '0123456789'
   

5. 调用百度接口识别我们截图的验证码图片

   def Disc_img():
       # 请求调用百度接口
   
       res = requests.get(
           'https://aip.baidubce.com/oauth/2.0/token?grant_type=client_credentials&client_id=edowjGnGuO1lTIAvmpmNUcrc&client_secret=Ob8txKjSChTdp2FvEMkdCBIwFoL0GKTx').json()
   
       token = res['access_token']
   
       # 识别图像
       url = 'https://aip.baidubce.com/rest/2.0/ocr/v1/accurate_basic?access_token=' + token
   
       # 构造头部
       headr = {'Contene-Type': 'application/x-www-form-urlencoded'}
   
       # 构造图片
   
       img = open('./register.png', 'rb')
   
       temp_img = img.read()
       # img.close()
   
       # 构建参数
       data = {'image': base64.b64encode(temp_img)}
       # 编码
       data = urllib.parse.urlencode(data)
   
       # 发送请求
   
       res = requests.post(url=url, data=data, headers=headr).text
       respone = json.loads(res)
       # 获取识别完毕的验证码
       code = respone['words_result'][0]['words']
       return code
   
   

6. 定位标签写入数据,把用户名和密码保存text

   def send_set():
       username = browser.find_element_by_xpath('//*[@id="app"]/div/section/div[2]/table/tr[1]/td[2]/input')
       username.clear()
       # 设定随机字符串
       user = ''
       # 随机生成用户名
       for i in range(6):
           num_index = random.randrange(len(num))
           num_s = num[num_index]
           user += num_s
       f = open('user.txt', 'a', encoding='utf-8')
       f.write('用户名:' + user)
       # 写入用户名
       username.send_keys(user)
       # 等待三秒
       time.sleep(2)
       # 定位密码框
       password = browser.find_element_by_xpath('//*[@id="app"]/div/section/div[2]/table/tr[2]/td[2]/input')
       password.clear()
       pwd = ''
       for i in range(6):
           num_index = random.randrange(len(num))
           num_s = num[num_index]
           pwd += num_s
       f = open('user.txt', 'a', encoding='utf-8')
       f.write('密码:' + pwd)
       password.send_keys(pwd)
       time.sleep(2)
       # 定位验证码输入框
       codem = browser.find_element_by_xpath('//*[@id="app"]/div/section/div[2]/table/tr[4]/td[2]/input')
       codem.clear()
       codem.send_keys(Disc_img())
       # 定位手机号
       phone = browser.find_element_by_xpath('//*[@id="app"]/div/section/div[2]/table/tr[5]/td[2]/input')
       phone.clear()
       # 随机手机号第二位
       start_two = str(random.choice([3, 4, 5, 7, 8]))
       end = ''
       for i in range(9):
           nums_index = random.randrange(len(nums))
           nums_s = nums[nums_index]
           end += nums_s
       phone.send_keys('1'+start_two+end)
       # 定位提交按钮
       but = browser.find_element_by_xpath('//*[@id="app"]/div/section/div[2]/table/tr[6]/td[2]/button')
       but.click()
       time.sleep(3)
       return
   
   
   if __name__ == '__main__':
       send_set()
   

7. 最终效果