Selenium利用Python图像识别解决验证码问题

Python图像识别参考：http://www.th7.cn/Program/Python/201602/768304.shtml

脚本思路：浏览器页面到达验证码输入页面，针对验证码图片进行另存为保存到本地，然后将图片复制到pytesser文件夹下（只有脚本和图片都在该文件夹下才能完成识别），然后调用Python图片识别处理验证码图片函数并返回验证码，然后就可以完成验证码输入操作！

运用到的模块：python图片识别、Python模拟键盘、Selenium鼠标事件

第一步：在C:\Python27\Lib\site-packages\pytesser路径下创建该Python文件，用作图片识别的方法

#coding=utf-8
'''
Created on 2018年4月22日

@author: Administrator
'''
from selenium import webdriver
import time,shutil,random,os
from selenium.webdriver.common.action_chains import ActionChains
import win32api,win32con
from PIL import Image
from io import BytesIO
from pytesser import *


def identifyImage(driver,element,imageName,browserPath='C:\\Users\\Administrator\\Downloads',pytesserPath='C:\\Python27\\Lib\\site-packages\\pytesser'):
    '''
    element:图片验证码元素;
    imageName:保存的验证码图片名称;
    browserPath:浏览器默认下载路径;
    pytesserPath:pytesser文件夹路径;
    DownloadPath:图片验证码另存为完成的文件路径（包含文件名）;
    savePath:图片验证码保存在pytesser中的路径（包含文件名）;
    '''
    DownloadPath = browserPath + '\\' + imageName
    savePath = pytesserPath + '\\' + imageName
    try:
        os.remove(DownloadPath)
    except WindowsError:
        pass
    try:
        os.remove(savePath)
    except WindowsError:
        pass
    action = ActionChains(driver)
    action.context_click(element).perform()
    win32api.keybd_event(86,0,0,0)
    win32api.keybd_event(86,0,win32con.KEYEVENTF_KEYUP,0)
    time.sleep(2)
    win32api.keybd_event(13,0,0,0)
    win32api.keybd_event(13,0,win32con.KEYEVENTF_KEYUP,0)
    time.sleep(2)
    shutil.copy(DownloadPath,pytesserPath)
    im = Image.open(savePath)
    imagetext = image_to_string(im).strip()
    return imagetext

第二步：实际场景调用图片识别

#coding=utf-8
'''
Created on 2018年4月22日

@author: Administrator
'''
from selenium import webdriver
import time,shutil,random,os
from selenium.webdriver.common.action_chains import ActionChains
import win32api,win32con
from pytesser.IdentifyImage import identifyImage
from PIL import Image
from io import BytesIO
from pytesser import *


def login(username=u'XXXXXX',password=u'XXXXXX'):   #登录函数
    driver= webdriver.Chrome()
    driver.maximize_window()
    driver.get('XXXXXX') 
    driver.implicitly_wait(30)
    driver.find_element_by_id('j_username').clear()
    driver.find_element_by_id('j_username').send_keys(username)
    driver.find_element_by_id('j_password').clear()
    driver.find_element_by_id('j_password').send_keys(password)
    driver.find_element_by_class_name('bigFont').click()
    time.sleep(2)
    try:
        driver.find_element_by_xpath('//span[contains(text(),"XXXXXX")]').click()
        driver.find_element_by_class_name('bigFont').click()
    except:
        pass
    return driver

def editpw(driver):    #用例函数
    driver.implicitly_wait(20)
    time.sleep(3)
    driver.find_element_by_class_name('icon-dept').click()
    try:
        driver.find_element_by_xpath('//div[@id="guide_buttons"]/img[2]').click()
    except Exception as e:
        print e
    elems = driver.find_elements_by_class_name('clientsearch-grid-text')
    random.choice(elems).click()
    driver.find_element_by_class_name('reSetPassword').click()
    time.sleep(2)
    elem = driver.find_element_by_class_name('verification-code-img')    #定位验证码图片
    value = identifyImage(driver,elem,'vCode.jpg')                        #调用Python图片识别方法
    driver.find_element_by_class_name('verification-code-text').send_keys(value)    #输入验证码
    time.sleep(2)
    driver.find_element_by_xpath('//span[contains(text(),"确认")]').click()
    try:
        message = driver.find_element_by_class_name('crm_prompt_text').text
        if message == u'密码已重置，请接收短信和邮件通知!':
            return True
        return False
    except:
        return False
    
driver = login()
Boole = editpw(driver)
print Boole
driver.quit()

PS：登录验证码无需通过该方法处理，可以直接通过添加Cookie来绕过登录验证码，然后该方法只能识别一些简单的图片验证码！

扫描二维码关注公众号，回复： 4312852 查看本文章