使用ddddocr实现文字验证码和滑块验证码

WRH一坨

已于 2024-05-20 10:06:27 修改

阅读量2.1k

点赞数 16

文章标签： python 自动化 selenium

于 2023-12-28 14:40:50 首次发布

本文链接：https://blog.csdn.net/weixin_43985642/article/details/135267930

版权

文字图片验证码识别

DDDDOCR YYDS！！！
ddddocr的数字英文识别准确率大概有个百分之九十，我用了挺久的，目前还没有遇到ddddocr识别准确率很低的数字英文验证码。

import ddddocr
# 识别验证码
def identifying_code(path):
    # show_ad广告
    ocr = ddddocr.DdddOcr(show_ad=False)
    with open(path, 'rb') as f:
        img_bytes = f.read()
    res = ocr.classification(img_bytes)
    print(res)
    return res

if __name__ == '__main__':
    result = identifying_code(r"C:\Users\Desktop\identifyingCode.jfif")

path放的是验证码图片的路径，会返回一个字符串，不过为了准确率，建议这边最好加一些判断，比如你的网站验证码是五个字符组成，如12ABC这样，就判断下字符串长度是否是5，如果不是直接刷新重新识别。

以下为滑块验证码代码

'''----------滑块验证码----------'''
import os
import base64
import time
import requests
from PIL import Image
from selenium import webdriver
from selenium.webdriver.common.by import By
from selenium.webdriver import ActionChains
from io import BytesIO
import traceback
import ddddocr
# 滑块图片路径
slider_path = r"C:\Users\Desktop\shuo\slider.png"
# 背景图片路径
background_path = r"C:\Users\Desktop\shuo\background.png"


def login(webpath):
    """geetest滑块验证码"""
    driver = webdriver.Chrome()
    driver.get(webpath)
    time.sleep(2)
    # 最大化窗口
    driver.maximize_window()
    time.sleep(2)
    driver.find_element(By.XPATH,
                        '/html/body/div[3]/div/div[4]/div[1]/form/div[2]/div/div/input').send_keys('S40433')
    time.sleep(1)
    driver.find_element(By.XPATH,
                        '/html/body/div[3]/div/div[4]/div[1]/form/div[3]/div/div/input').send_keys('misumi@02')
    time.sleep(2)
    driver.find_element(By.XPATH, '/html/body/div[3]/div/div[4]/div[1]/form/button').click()
    time.sleep(5)

    for i in range(8):
        slider_url = driver.find_element(By.XPATH, '/html/body/div[3]/div/div[4]/div[2]/div/div[2]/div/div[1]/div/img').get_attribute("src")
        slider_url = slider_url.replace('data:image/png;base64,','')
        print(slider_url)
        background_url = driver.find_element(By.XPATH, '/html/body/div[3]/div/div[4]/div[2]/div/div[2]/div/div[2]/div/div/div/img').get_attribute("src")
        background_url = background_url.replace('data:image/png;base64,','')
        print(background_url)
        time.sleep(2)
        base64_to_image(slider_url,slider_path)
        base64_to_image(background_url,background_path)
        # 下载图片
        # req_slider = requests.get(slider_url)
        # with open(r"C:\Users\Desktop\shuo\slider.png", "wb") as f:
        #     f.write(req_slider.content)
        # req_background = requests.get(background_url)
        # with open(r"C:\Users\Desktop\shuo\background.png", "wb") as f:
        #     f.write(req_background.content)

        # 获取滑块在背景图片的位置
        det = ddddocr.DdddOcr(det=False, ocr=False, show_ad=False)
        with open(slider_path, 'rb') as f:
            target_bytes = f.read()
        with open(background_path, 'rb') as f:
            background_bytes = f.read()
        res = det.slide_match(target_bytes=target_bytes, background_bytes=background_bytes, simple_target=True)
        print(res)
        # 计算滑块轨迹
        distance = res["target"][0]-10
        # tracks = [distance]
        tracks = get_tracks(distance)
        print(tracks)
        time.sleep(2)
        slider_button = driver.find_element(By.XPATH, "/html/body/div[3]/div/div[4]/div[2]/div/div[2]/div/div[2]/div/div/div/img")
        # 移动滑块
        move_to_gap(driver, slider_button, tracks)
        time.sleep(2)

    driver.close()



def test():
    # 测试 滑块到背景之间得距离
    det = ddddocr.DdddOcr(det=False, ocr=False, show_ad=False)
    with open(slider_path, 'rb') as f:
        target_bytes = f.read()
    with open(background_path, 'rb') as f:
        background_bytes = f.read()
    res = det.slide_match(target_bytes=target_bytes, background_bytes=background_bytes, simple_target=True)
    print(res)


def transPNG(srcImageName, dstImageName):
    '''
    图片透明化处理
    :param srcImageName: 原图片路径
    :param dstImageName: 处理完的图片路径
    :return:
    '''
    img = Image.open(srcImageName)
    img = img.convert("RGBA")
    datas = img.getdata()
    newData = list()
    for item in datas:
        if item[0] > 220 and item[1] > 220 and item[2] > 220:
            newData.append((255, 255, 255, 0))
        else:
            newData.append(item)
    img.putdata(newData)
    img.save(dstImageName, "PNG")


def get_tracks(distance):
    """
    根据偏移量获取移动轨迹
    :param distance:偏移量
    :return:移动轨迹
    """
    # 移动轨迹
    tracks = []
    # 减速阈值
    mid = distance * 4 / 5

    for i in range(5):
        tracks.append(mid/5)
    for j in range(2):
        tracks.append(distance/5/2)

    return tracks


def move_to_gap(browser, slider, tracks):
    """
    拖动滑块
    :param browser 浏览器
    :param slider: 滑块
    :param tracks: 轨迹
    :return:
    """
    # 模拟滑动滑块
    action = ActionChains(browser)
    action.click_and_hold(slider).perform()
    # action.reset_actions()   # 清除之前的action
    for i in tracks:
        action.move_by_offset(xoffset=i, yoffset=0).perform()
    time.sleep(0.5)
    action.release().perform()

def base64_to_image(base64_string,save_path):
    try:
        image_data = base64.b64decode(base64_string)
        image = Image.open(BytesIO(image_data))
        image.save(save_path)
    except:
        print(traceback.format_exc())
    # return image
if __name__ == '__main__':
	web_path = ""
    login(web_path)