pip install selenium

#圖像處理標(biāo)準(zhǔn)庫(kù)
from PIL import Image  
#web測(cè)試
from selenium import webdriver
#鼠標(biāo)操作
from selenium.webdriver.common.action_chains import ActionChains
#等待時(shí)間 產(chǎn)生隨機(jī)數(shù) 
import time,random

#打開頁面至屏幕最大尺寸
driver = webdriver.Chrome()
driver.get('https://account.ch.com/NonRegistrations-Regist')
driver.maximize_window()
#獲取輸入手機(jī)號(hào)碼的表單
input1 = driver.find_element_by_name('phoneNumberInput')
# 輸入注冊(cè)號(hào)碼
input1.send_keys(phoneNumber)
time.sleep(0.2)
#獲取打開滑塊驗(yàn)證碼頁面的元素
getcheck=driver.find_element_by_id('getDynamicPwd')
#點(diǎn)擊進(jìn)入滑塊驗(yàn)證碼頁面
getcheck.click()

# 獲取拖拽的圓球
slideblock = driver.find_element_by_class_name('geetest_slider_button')
# 鼠標(biāo)點(diǎn)擊圓球不松開
ActionChains(driver).click_and_hold(slideblock).perform()
# 將圓球滑至相對(duì)起點(diǎn)位置的最右邊
ActionChains(driver).move_by_offset(xoffset=250, yoffset=0).perform()
time.sleep(0.4)
# 保存包含滑塊及缺口的頁面截圖
driver.save_screenshot('D:\quekou.png')
# 放開圓球
ActionChains(driver).release(slideblock).perform()
#打開保存至本地的缺口頁面截圖
quekouimg=Image.open('d://quekou.png')
# 匹配本地對(duì)應(yīng)原圖
sourceimg=match_source(quekouimg)
def match_source(image):
  imagea=Image.open('d://source1.png')
  imageb=Image.open('d://source2.png')
  imagec=Image.open('d://source3.png')
  imaged=Image.open('d://source4.png')
  list=[imagea,imageb,imagec,imaged]
  #通過像素差遍歷匹配本地原圖
  for i in list:
    #本人電腦原圖與缺口圖對(duì)應(yīng)滑塊圖片橫坐標(biāo)相同，縱坐標(biāo)原圖比缺口圖大88px，可根據(jù)實(shí)際情況修改
    pixel1=image.getpixel((868,340))
    pixel2=i.getpixel((868,428))
    #pixel[0]代表R值，pixel[1]代表G值，pixel[2]代表B值
    if abs(pixel1[0]-pixel2[0])5:
      return i
  return image

# 獲取缺口位置
visualstack=get_diff_location(sourceimg,quekouimg)
# 獲取移動(dòng)距離loc，827為滑塊起點(diǎn)位置
loc=visualstack-827
# 計(jì)算滑塊位移距離
def get_diff_location(image1,image2):
  #（825,1082）（335,463）為滑塊圖片區(qū)域，可根據(jù)實(shí)際情況修改
  for i in range(825,1082):
    for j in range(335,463):
      #遍歷原圖與缺口圖像素值尋找缺口位置
      if is_similar(image1,image2,i,j)==False:
        return i
  return -1
# 對(duì)比RGB值得到缺口位置
def is_similar(image1,image2,x,y):
  pixel1=image1.getpixel((x, y+88))
  pixel2=image2.getpixel((x, y))
  # 截圖像素也許存在誤差，50作為容差范圍
  if abs(pixel1[0]-pixel2[0])>=50 and abs(pixel1[1]-pixel2[1])>=50 and abs(pixel1[2]-pixel2[2])>=50:
    return False
  return True

#滑塊移動(dòng)軌跡
def get_track(self,distance):
  track=[]
  current=0
  mid=distance*3/4
  t=random.randint(2,3)/10
  v=0
  while currentdistance:
     if currentmid:
       a=2
     else:
       a=-3
     v0=v
     v=v0+a*t
     move=v0*t+1/2*a*t*t
     current+=move
     track.append(round(move))
  return track
# 生成拖拽移動(dòng)軌跡，加3是為了模擬滑過缺口位置后返回缺口的情況
track_list=get_track(loc+3)
time.sleep(2)
ActionChains(driver).click_and_hold(slideblock).perform()
time.sleep(0.2)
# 根據(jù)軌跡拖拽圓球
for track in track_list:
  ActionChains(driver).move_by_offset(xoffset=track,yoffset=0).perform()
# 模擬人工滑動(dòng)超過缺口位置返回至缺口的情況，數(shù)據(jù)來源于人工滑動(dòng)軌跡，同時(shí)還加入了隨機(jī)數(shù)，都是為了更貼近人工滑動(dòng)軌跡
imitate=ActionChains(driver).move_by_offset(xoffset=-1, yoffset=0)
time.sleep(0.015)
imitate.perform()
time.sleep(random.randint(6,10)/10)
imitate.perform()
time.sleep(0.04)
imitate.perform()
time.sleep(0.012)
imitate.perform()
time.sleep(0.019)
imitate.perform()
time.sleep(0.033)
ActionChains(driver).move_by_offset(xoffset=1, yoffset=0).perform()
# 放開圓球
ActionChains(driver).pause(random.randint(6,14)/10).release(slideblock).perform()
time.sleep(2)
#務(wù)必記得加入quit()或close()結(jié)束進(jìn)程，不斷測(cè)試電腦只會(huì)卡卡西
driver.close()