一、簡(jiǎn)介

初次開始使用Selenium，于是開始用登錄作為聯(lián)手項(xiàng)目，是一個(gè)真實(shí)的登錄界面，包含驗(yàn)證碼（驗(yàn)證碼是難點(diǎn)，獲取與識(shí)別），以下會(huì)講的很詳細(xì)，我怕我下次又忘記了
我采用Selenium + python + 百度云OCR，安裝等教程自行百度，我就不多說(shuō)了

二、selenium打開瀏覽器獲取相應(yīng)el

from selenium import webdriver
from selenium.webdriver.common.by import By
import base64
#注冊(cè)Chrome驅(qū)動(dòng),打開瀏覽器網(wǎng)址
driver = webdriver.Chrome()
url = "************"
driver.get(url)
#獲取相應(yīng)el
driver.find_element(By.NAME,"***").send_keys('***')
driver.find_element(By.NAME, "***").send_keys('***')
#獲取圖片
png = driver.find_element(By.CSS_SELECTOR,"****")
img_url = png.get_attribute("src")  # 獲得圖片的url，本次測(cè)試的是bs64生成，采用解碼獲得圖片
head, context = img_url.split(",")  # 截取后字段的bas64編碼
img_data = base64.b64decode(context)
image = Image.open(BytesIO(img_data));
image.save("capt.png")

這段代碼主要是利用selenium打開網(wǎng)站，獲取到相應(yīng)的dom，find_element_by_id等寫法已經(jīng)統(tǒng)一換為find_element(By.ID, “***”)等了，先將圖片保存下來(lái)考研方便后續(xù)的識(shí)別

三、驗(yàn)證碼預(yù)處理

驗(yàn)證碼一般無(wú)法直接識(shí)別，我們需要通過(guò)灰度、二值、降噪等手段，使得圖片更加清晰，提高圖片的識(shí)別準(zhǔn)確率
下列代碼需要提前引入以下包

from PIL import Image

1.灰度化

img = img.convert('L')  # P模式轉(zhuǎn)換為L(zhǎng)模式(灰度模式默認(rèn)閾值127)

2.二值化

count = 230  # 設(shè)定閾值，閾值的大小由圖片的清晰度等自行調(diào)節(jié)
table = []
for i in range(256):
    if i < count:
        table.append(0)
    else:
        table.append(1)
img = img.point(table, '1')
img.save('captcha.png')

3.降噪

pixdata = img.load()
w,h = img.size
for y in range(1,h-1):
   for x in range(1,w-1):
        count = 0
        if pixdata[x,y-1] > 245:
            count = count + 1
        if pixdata[x,y+1] > 245:
            count = count + 1
        if pixdata[x-1,y] > 245:
            count = count + 1
        if pixdata[x+1,y] > 245:
            count = count + 1
        if count > 2:
            pixdata[x,y] = 255
return img

四、百度云識(shí)別

需要提前注冊(cè)文章來(lái)源地址http://www.zghlxwxcb.cn/news/detail-788427.html

from aip import AipOcr
# 識(shí)別碼
APP_ID = '****'
API_KEY = '****'
SECRET_KEY = '****'
# 初始化對(duì)象
client = AipOcr(APP_ID, API_KEY, SECRET_KEY)
# 讀取圖片
def get_file_content(file_path):
    with open(file_path, 'rb') as f:
        return f.read()

image = get_file_content('end.png') # 傳入圖片
# 定義參數(shù)變量
options = {'language_type': 'ENG', }  # 識(shí)別語(yǔ)言類型，默認(rèn)為'CHN_ENG'中英文混合
#  調(diào)用通用文字識(shí)別
result = client.basicGeneral(image, options)  # 高精度接口 basicAccurate
for word in result['words_result']:
    captcha = (word['words'])

print('識(shí)別結(jié)果：' + captcha)
return captcha

五、完整代碼

import base64
from io import BytesIO
from selenium import webdriver
from selenium.webdriver.common.by import By
from PIL import Image
from time import sleep
from aip import AipOcr

# 對(duì)二值化的圖像進(jìn)行降噪處理，使得識(shí)別更加準(zhǔn)確
def depoint(img):
    pixdata = img.load()
    w,h = img.size
    for y in range(1,h-1):
        for x in range(1,w-1):
            count = 0
            if pixdata[x,y-1] > 245:
                count = count + 1
            if pixdata[x,y+1] > 245:
                count = count + 1
            if pixdata[x-1,y] > 245:
                count = count + 1
            if pixdata[x+1,y] > 245:
                count = count + 1
            if count > 2:
                pixdata[x,y] = 255
    return img

# 獲得驗(yàn)證碼，并進(jìn)行處理，二值化
def deal_image():
    png = driver.find_element(By.CSS_SELECTOR,".*******")
    img_url = png.get_attribute("src")  # 獲得圖片的url，本次測(cè)試的是bs64生成，采用解碼獲得圖片
    head, context = img_url.split(",")  # 截取后字段的bas64編碼
    img_data = base64.b64decode(context)
    image = Image.open(BytesIO(img_data));
    image.save("capt.png")

    img = Image.open('capt.png')
    img = img.convert('L')  # P模式轉(zhuǎn)換為L(zhǎng)模式(灰度模式默認(rèn)閾值127)
    count = 230  # 設(shè)定閾值，閾值的大小由圖片的清晰度等自行調(diào)節(jié)
    table = []
    for i in range(256):
        if i < count:
            table.append(0)
        else:
            table.append(1)

    # print(table)

    img = img.point(table, '1')
    img.save('captcha.png')

# 調(diào)用百度云OCR API 進(jìn)行識(shí)別
def discern_captcha():
    # 識(shí)別碼
    APP_ID = '****'
    API_KEY = '****'
    SECRET_KEY = '****'
    # 初始化對(duì)象
    client = AipOcr(APP_ID, API_KEY, SECRET_KEY)
    # 讀取圖片
    def get_file_content(file_path):
        with open(file_path, 'rb') as f:
            return f.read()

    image = get_file_content('end.png') # 傳入圖片
    # 定義參數(shù)變量
    options = {'language_type': 'ENG', }  # 識(shí)別語(yǔ)言類型，默認(rèn)為'CHN_ENG'中英文混合
    #  調(diào)用通用文字識(shí)別
    result = client.basicGeneral(image, options)  # 高精度接口 basicAccurate
    for word in result['words_result']:
        captcha = (word['words'])

    print('識(shí)別結(jié)果：' + captcha)
    return captcha

# 登錄事件，若識(shí)別不準(zhǔn)確需點(diǎn)擊第二次重復(fù)
def ClickRandomCode():
    deal_image()
    image = depoint(Image.open('captcha.png'))
    image.save('end.png')
    image = depoint(Image.open('end.png'))
    image.save("end.png")
    randomCode = discern_captcha()
    driver.find_element(By.NAME, "randomCode").send_keys(randomCode)
    driver.find_element(By.CSS_SELECTOR, ".card-z .el-button").click()
    sleep(2) # 錯(cuò)誤或成功信息有延遲
    try:
        text = driver.find_element(By.CSS_SELECTOR, ".el-message__content").text
        if (text == "驗(yàn)證碼錯(cuò)誤"):
            driver.find_element(By.CSS_SELECTOR, ".verification-code > img").click()
            driver.find_element(By.CSS_SELECTOR, ".verification-code > img").clear()
            ClickRandomCode() # 遞歸調(diào)用直至正確
            print("驗(yàn)證碼錯(cuò)誤")
        else:
            print("登錄成功")
    except:
        print("ok")

def login(***,***):
    driver.find_element(By.NAME,"****").send_keys(***)
    driver.find_element(By.NAME, "***").send_keys(***)
    ClickRandomCode()

if __name__ == '__main__':
    driver = webdriver.Chrome()
    url = "*****"
    driver.get(url)
    login("****","****")

到了這里，關(guān)于Selenium 初步使用之自動(dòng)登錄（包含識(shí)別驗(yàn)證碼），超級(jí)詳細(xì)的文章就介紹完了。如果您還想了解更多內(nèi)容，請(qǐng)?jiān)谟疑辖撬阉鱐OY模板網(wǎng)以前的文章或繼續(xù)瀏覽下面的相關(guān)文章，希望大家以后多多支持TOY模板網(wǎng)！