当前位置 博文首页 > python自动化操作之动态验证码、滑动验证码的降噪和识别
python对动态验证码、滑动验证码的降噪和识别，在各种自动化操作中，我们经常要遇到沿跳过验证码的操作，而对于验证码的降噪和识别，的确困然了很多的人。这里我们就详细讲解一下不同验证码的降噪和识别。
from selenium import webdriver
from PIL import Image
# 实例化浏览器
driver = webdriver.Chrome()
# 最大化窗口
driver.maximize_window()
# 打开登陆页面
driver.get(# 你的url地址)
# 保存页面截图
driver.get_screenshot_as_file('./screen.png')
# 定位验证码的位置
location = driver.find_element_by_id('login_yzm_img').location
size = driver.find_element_by_id('login_yzm_img').size
left = location['x']
top = location['y']
right = location['x'] + size['width']
bottom = location['y'] + size['height']
# 裁剪保存
img = Image.open('./screen.png').crop((left,top,right,bottom))
img.save('./code.png')
driver.quit()
# 下载两个滑块
bg = self.driver.find_element_by_xpath('//*[@]/div/div[2]/img[1]').get_attribute('src')
slider = self.driver.find_element_by_xpath('//*[@]/div/div[2]/img[2]').get_attribute('src')
request.urlretrieve(bg, os.getcwd() + '/bg.png')
request.urlretrieve(slider, os.getcwd() + '/slider.png')
# 获取两个滑块偏移量方法
def getGap(self, sliderImage, bgImage):
'''
Get the gap distance
:param sliderImage: the image of slider
:param bgImage: the image of background
:return: int
'''
bgImageInfo = cv2.imread(bgImage, 0)
bgWidth, bgHeight = bgImageInfo.shape[::-1]
bgRgb = cv2.imread(bgImage)
bgGray = cv2.cvtColor(bgRgb, cv2.COLOR_BGR2GRAY)
slider = cv2.imread(sliderImage, 0)
sliderWidth, sliderHeight = slider.shape[::-1]
res = cv2.matchTemplate(bgGray, slider, cv2.TM_CCOEFF)
a, b, c, d = cv2.minMaxLoc(res)
# print(a,b,c,d)
# 正常如下即可
# return c[0] if abs(a) >= abs(b) else d[0]
# 但是头条显示验证码的框跟验证码本身的像素不一致，所以需要根据比例计算
if abs(a) >= abs(b):
return c[0] * bgWidth / (bgWidth - sliderWidth)
else:
return d[0] * bgWidth / (bgWidth - sliderWidth)
# 移动路径方法
def getTrack(self, distance):
'''
Get the track by the distance
:param distance: the distance of gap
:return: list
'''
# 移动轨迹
track = []
# 当前位移
current = 0
# 减速阈值
mid = distance * 4 / 5
# 计算间隔
t = 0.2
# 初速度
v = 0
while current < distance:
if current < mid:
# 加速度为正2
a = 2
else:
# 加速度为负3
a = -3
# 初速度v0
v0 = v
# 当前速度v = v0 + at
v = v0 + a * t
# 移动距离x = v0t + 1/2 * a * t^2
move = v0 * t + 1 / 2 * a * t * t
# 当前位移
current += move
# 加入轨迹
track.append(round(move))
return track
# 滑动到缺口位置
def moveToGap(self, track):
'''
Drag the mouse to gap
:param track: the track of mouse
:return: None
'''
ActionChains(self.driver).click_and_hold(self.driver.find_element_by_xpath('//*[@]/div/div[3]/div[2]/div[2]/div')).perform()
while track:
x = random.choice(track)
ActionChains(self.driver).move_by_offset(xoffset=x, yoffset=0).perform()
track.remove(x)
time.sleep(0.5)
ActionChains(self.driver).release().perform()
验证码的降噪，只是为了处理验证码图像上的多余的线条和干扰线，让你后期识别更加的准确，提高识别的准确度
第一步：可以进行灰度转化
img = cv2.imread('yzm.png')
# 将图片灰度化处理,降维,加权进行灰度化c
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
cv2.imshow('min_gray',gray)
cv2.waitKey(0)
cv2.destroyAllWindows()
第二步： 二值化处理
import cv2
img = cv2.imread('yzm.png')
# 将图片灰度化处理,降维,加权进行灰度化c
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
t,gray2 = cv2.threshold(gray,220,255,cv2.THRESH_BINARY)
cv2.imshow('threshold',gray2)
cv2.waitKey(0)
cv2.destroyAllWindows()
第三步：噪点过滤
import cv2
img = cv2.imread('yzm.png')
# 将图片灰度化处理,降维,加权进行灰度化c
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
t,gray2 = cv2.threshold(gray,220,255,cv2.THRESH_BINARY)
def remove_noise(img, k=4):
img2 = img.copy()
# img处理数据，k过滤条件
w, h = img2.shape
def get_neighbors(img3, r, c):
count = 0
for i in [r - 1, r, r + 1]:
for j in [c - 1, c, c + 1]:
if img3[i, j] > 10: # 纯白色
count += 1
return count
# 两层for循环判断所有的点
for x in range(w):
for y in range(h):
if x == 0 or y == 0 or x == w - 1 or y == h - 1:
img2[x, y] = 255
else:
n = get_neighbors(img2, x, y) # 获取邻居数量，纯白色的邻居
if n > k:
img2[x, y] = 255
return img2
result = remove_noise(gray2)
cv2.imshow('8neighbors', result)
cv2.waitKey(0)
cv2.destroyAllWindows()
通常我们会使用tesserocr识别验证码，但是这个库有很大的局限性，识别率低，即时降噪效果很好，有很少的线条，也会不准确，这种识别方式并不十分推荐
所以我们一般会使用第三方的接口进行识别，比如阿里的图片识别、腾讯也都是有的
这些第三方接口需要自己接入识别接口
#识别降噪后的图片 code = tesserocr.image_to_text(nrImg) #消除空白字符 code.strip() #打印 print(code)
