python3 使用OpenCV计算滑块拼图验证码缺口位置(场景示例)

import cv2

# 作者-上海悠悠 QQ交流群:717225969
# blog地址 https://www.cnblogs.com/yoyoketang/

def show(name):
    '''展示圈出来的位置'''
    cv2.imshow('Show', name)
    cv2.waitKey(0)
    cv2.destroyAllWindows()


def _tran_canny(image):
    """消除噪声"""
    image = cv2.GaussianBlur(image, (3, 3), 0)
    return cv2.Canny(image, 50, 150)


def detect_displacement(img_slider_path, image_background_path):
    """detect displacement"""
    # # 参数0是灰度模式
    image = cv2.imread(img_slider_path, 0)
    template = cv2.imread(image_background_path, 0)

    # 寻找最佳匹配
    res = cv2.matchTemplate(_tran_canny(image), _tran_canny(template), cv2.TM_CCOEFF_NORMED)
    # 最小值，最大值，并得到最小值, 最大值的索引
    min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(res)

    top_left = max_loc[0]  # 横坐标
    # 展示圈出来的区域
    x, y = max_loc  # 获取x,y位置坐标

    w, h = image.shape[::-1]  # 宽高
    cv2.rectangle(template, (x, y), (x + w, y + h), (7, 249, 151), 2)
    show(template)
    return top_left

if __name__ == '__main__':
    top_left = detect_displacement("target.png", "background.png")
    print(top_left)

# 展示圈出来的区域
  #  x, y = max_loc  # 获取x,y位置坐标

  #  w, h = image.shape[::-1]  # 宽高
  #  cv2.rectangle(template, (x, y), (x + w, y + h), (7, 249, 151), 2)
  #  show(template)

#coding=utf-8
import cv2
import numpy as np
from PIL import Image
 
def get_element_slide_distance():
 
 otemp = 'captcha2.png'
 oblk = 'captcha1.png'
 target = cv2.imread(otemp, 0) # 读取进行色度图片，转换为numpy中的数组类型数据
 template = cv2.imread(oblk, 0)
 width, height = target.shape[::-1] # 获取缺口图数组的形状 -->缺口图的宽和高
 temp = 'temp.jpg' # 将处理之后的图片另存
 targ = 'targ.jpg'
 cv2.imwrite(temp, template)
 cv2.imwrite(targ, target)
 target = cv2.imread(targ) # 读取另存的滑块图
 target = cv2.cvtColor(target, cv2.COLOR_BGR2GRAY) # 进行色彩转换
 # 去除白色部分 获取滑块正常大小
 target = target[target.any(1)]
 
 target = abs(255 - target) # 获取色差的绝对值
 cv2.imwrite(targ, target) # 保存图片
 target = cv2.imread(targ) # 读取滑块
 template = cv2.imread(temp) # 读取背景图
 result = cv2.matchTemplate(target, template, cv2.TM_CCOEFF_NORMED) # 比较两张图的重叠区域
 top, left = np.unravel_index(result.argmax(), result.shape) # 获取图片的缺口位置
 #缺口位置
 print((left, top, left + width, top + height)) # 背景图中的图片缺口坐标位置
 
 #调用PIL Image 做测试
 image = Image.open("captcha1.png")
 
 rectangle = (left + 3, top + 3, left + width - 3, top + height - 3) #去掉白色块的影响(上面去掉白色部分的功能并没有真的起作用)
 #切割
 imagecrop = image.crop(rectangle)
 #保存切割的缺口
 imagecrop.save("new_image.jpg")
 
 return left+3
 
distance = get_element_slide_distance()
# 滑动距离误差校正，滑动距离*图片在网页上显示的缩放比-滑块相对的初始位置
distance = distance*(280/680) - 22

def generate_tracks1(XCoordinates):
 element = browser.find_element_by_xpath("//div[@class='secsdk-captcha-drag-icon sc-jKJlTe fsBatO']")
 ActionChains(browser).click_and_hold(on_element = element).perform()
 #
 # ActionChains(browser).move_by_offset(xoffset=0, yoffset=y - 445).perform()
 #
 # time.sleep(0.15)
 # print("第二步，拖动元素")
 distance = XCoordinates - 60
 # 初速度
 v = 0
 # 单位时间为0.2s来统计轨迹，轨迹即0.2内的位移,越低看起来越丝滑！！
 t = 0.08
 # 位移/轨迹列表，列表内的一个元素代表0.2s的位移
 tracks = []
 # 当前的位移
 current = 0
 # 到达mid值开始减速
 mid = distance * 5 / 8
 
 distance += 10  # 先滑过一点，最后再反着滑动回来
 # a = random.randint(1,3)
 while current < distance:
  if current < mid:
   # 加速度越小，单位时间的位移越小,模拟的轨迹就越多越详细
   a = random.randint(100, 200)  # 加速运动
  else:
   a = -random.randint(2, 10)  # 减速运动
 
  # 初速度
  v0 = v
  # 0.2秒时间内的位移
  s = v0 * t + 0.5 * a * (t ** 2)
  # 当前的位置
  current += s
  # 添加到轨迹列表
  tracks.append(round(s))
 
  # 速度已经达到v,该速度作为下次的初速度
  v = v0 + a * t
  if current > distance:
   break
 
 # 反着滑动到大概准确位置
 # for i in range(4):
 #     tracks.append(-random.randint(1, 3))
 # for i in range(4):
 #    tracks.append(-random.randint(1,3))
 random.shuffle(tracks)
 count = 0
 for item in tracks:
  print(item)
  count += item
  ActionChains(browser).move_by_offset(xoffset = item, yoffset = random.randint(-2, 2)).perform()
 
 # ActionChains(browser).move_to_element_with_offset(to_element=element, xoffset=XCoordinates-18,yoffset=y - 445).perform()
 # time.sleep(2)
 # # 释放鼠标
 print(count)
 ActionChains(browser).release(on_element = element).perform()