【APP逆向38】滑块处理

• 当我们在逆向时，有的登录界面，需要滑块或者点选之类的图片验证
• 这篇我们主要了解使用opencv识别滑块
• 1.获取滑块图片并且保存在本地

import base64
import requestsres = requests.post(url='https://m.captcha.qq.com/',data='{"Action":"CheckCaptchaAppId_v1.0.1","CaptchaAppId":2075530305}',
)
esid = res.json()['Response']['SId']res = requests.post(url='https://m.captcha.qq.com/',data='{"Action":"GetImageData_v1.0.1","ESId":"%s"}' % esid,
)
data_dict = res.json()# 背景图片，base64
l_image = data_dict['Response']['ImageDataL']
l_image_content = base64.b64decode(bytes(l_image, encoding="ascii"))
with open('l_image.png', mode='wb') as f:f.write(l_image_content)# 缺口图片，base64
s_image = data_dict['Response']['ImageDataS']
s_image_content = base64.b64decode(bytes(s_image, encoding="ascii"))
with open('s_image.png', mode='wb') as f:f.write(s_image_content)

• 2.识别并且计算距离

import base64
import requests
import jsonimport cv2
import numpy as npdef get_image_object(bs_image):obj = base64.b64decode(bytes(bs_image, encoding="ascii"))img_buffer_np = np.frombuffer(obj, dtype=np.uint8)# 从指定的内存缓存中读取一维numpy数据, 并把数据转换(解码)成图像矩阵格式img_np = cv2.imdecode(img_buffer_np, 1)bg_img = cv2.cvtColor(img_np, cv2.COLOR_BGR2GRAY)return bg_imgdef get_distance(l_image_object, s_image_object):# 边缘检测bg_edge = cv2.Canny(l_image_object, 255, 255)tp_edge = cv2.Canny(s_image_object, 255, 255)bg_pic = cv2.cvtColor(bg_edge, cv2.COLOR_GRAY2RGB)tp_pic = cv2.cvtColor(tp_edge, cv2.COLOR_GRAY2RGB)res = cv2.matchTemplate(bg_pic, tp_pic, cv2.TM_CCOEFF_NORMED)min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(res)  # 寻找最优匹配x = max_loc[0]return xres = requests.post(url='https://m.captcha.qq.com/',data='{"Action":"CheckCaptchaAppId_v1.0.1","CaptchaAppId":2075530305}',
)
esid = res.json()['Response']['SId']res = requests.post(url='https://m.captcha.qq.com/',data='{"Action":"GetImageData_v1.0.1","ESId":"%s"}' % esid,
)
data_dict = res.json()l_image = data_dict['Response']['ImageDataL']
l_image_object = get_image_object(l_image)s_image = data_dict['Response']['ImageDataS']
s_image_object = get_image_object(s_image)# 计算距离
distance = get_distance(l_image_object, s_image_object)
print(distance)

• 3.提交滑块

import base64
import requests
import json
import time
import random
import cv2
import numpy as npdef get_image_object(bs_image):obj = base64.b64decode(bytes(bs_image, encoding="ascii"))img_buffer_np = np.frombuffer(obj, dtype=np.uint8)# 从指定的内存缓存中读取一维numpy数据, 并把数据转换(解码)成图像矩阵格式img_np = cv2.imdecode(img_buffer_np, 1)bg_img = cv2.cvtColor(img_np, cv2.COLOR_BGR2GRAY)return bg_imgdef get_distance(l_image_object, s_image_object):# 边缘检测bg_edge = cv2.Canny(l_image_object, 255, 255)tp_edge = cv2.Canny(s_image_object, 255, 255)bg_pic = cv2.cvtColor(bg_edge, cv2.COLOR_GRAY2RGB)tp_pic = cv2.cvtColor(tp_edge, cv2.COLOR_GRAY2RGB)res = cv2.matchTemplate(bg_pic, tp_pic, cv2.TM_CCOEFF_NORMED)min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(res)  # 寻找最优匹配x = max_loc[0]return xres = requests.post(url='https://m.captcha.qq.com/',data='{"Action":"CheckCaptchaAppId_v1.0.1","CaptchaAppId":2075530305}',
)
esid = res.json()['Response']['SId']res = requests.post(url='https://m.captcha.qq.com/',data='{"Action":"GetImageData_v1.0.1","ESId":"%s"}' % esid,
)
data_dict = res.json()l_image = data_dict['Response']['ImageDataL']
l_image_object = get_image_object(l_image)s_image = data_dict['Response']['ImageDataS']
s_image_object = get_image_object(s_image)
distance = get_distance(l_image_object, s_image_object)sid = data_dict['Response']['UniqueSId']# 提交滑块
time.sleep(random.randint(3, 5))
data = '{"Action":"VerificationCaptchaImageAnswer_v1.0.1","UniqueSId":"%s","LeftTopX":"%s","ranNum":0,"Frequency":1}' % (sid, distance)
res = requests.post(url='https://m.captcha.qq.com/',data=data,
)print(res.text)