使用pytesseract实现自动验证码

对图片进行二值化处理

import pytesseract
from PIL import Image

def shibie(image):
    # 打开图片
    img = Image.open(image)
    # 图片转化为灰度图片
    img = img.convert('L')

    # 二值化处理
    threshold = 140
    table = []
    for i in range(256):
        if i < threshold:
            table.append(0)
        else:
            table.append(1)
    out = img.point(table, '1')

    img = img.convert('RGB')

    return pytesseract.image_to_string(img)

实现代码

import requests
from bs4 import BeautifulSoup
import time

# 创建一个会话
from shibie import shibie

s = requests.Session()

i = 1
while 1:
    # 将验证码写入到本地
    get_url = 'https://so.gushiwen.org/user/login.aspx?from=http://so.gushiwen.org/user/collect.aspx'
    headers = {
        'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.108 Safari/537.36',
    }
    r_get = s.get(url=get_url,headers=headers)
    # 通过bs4得到图片
    soup = BeautifulSoup(r_get.text, 'lxml')
    image_src = 'https://so.gushiwen.org' + soup.find('img',id='imgCode')['src']
    # 通过会话进行请求
    r_img = s.get(image_src)
    # 将图片写到文件中
    with open('code.png','wb') as fq:
        fq.write(r_img.content)

    # 获取两个隐藏值
    viewstate = soup.find('input', id='__VIEWSTATE')['value']
    viewg = soup.find('input',id='__VIEWSTATEGENERATOR')['value']

    post_url = 'https://so.gushiwen.org/user/login.aspx?from=http%3a%2f%2fso.gushiwen.org%2fuser%2fcollect.aspx'

    # 调用函数 自动识别
    code = shibie('code.png')

    data = {
        '__VIEWSTATE': viewstate,
        '__VIEWSTATEGENERATOR': viewg,
        'from': 'http://so.gushiwen.org/user/collect.aspx',
        'email': '123456@qq.com',
        'pwd': '123456',
        'code': code,
        'denglu': '登录',
    }

    r_post = s.post(url=post_url,headers=headers,data=data)

    # 判断是否登陆成功
    if '退出登录' in r_post.text:
        print('登陆成功')
        break

    print('第%s次登陆失败' % i)
    i += 1
    time.sleep(2)