验证码：EasyDL 机器学习识别与云码平台一站式识别
目录
EasyDL 机器学习识别（实践：京东商城）
（一）批量获取验证码图片
（二）EasyDL机器学习（百度智能云）
（三）调用EasyDLAPI接口识别验证码
云码平台一站式识别（实战：豆瓣网）
（一）获取验证码图片
（二）验证码识别
（三）动态获取验证码
EasyDL 机器学习识别（实践：京东商城）
（一）批量获取验证码图片
from selenium import webdriver
from selenium.webdriver.chrome.service import Service
from selenium.webdriver.common.by import By
from selenium.webdriver import ActionChains
import time
import requests
service = Service(executable_path="../_resources/chromedriver")
driver = webdriver.Chrome(service=service)
driver.get("https://passport.jd.com/uc/login")
select = driver.find_element(By.XPATH, '/html/body/div[2]/div[2]/div[1]/div/div[3]/a')
select.click()
username = driver.find_element(By.XPATH, '//*[@id="loginname"]')
ActionChains(driver).pause(0.5).click(username).send_keys("xxxxxxxxxx").parform()
password = driver.find_element(By.XPATH, '//*[@id="nloginpwd"]')
ActionChains(driver).pause(0.5).click(password).send_keys("xxxxxxxxxx").parform()
submit = driver.find_element(By.XPATH, '//*[@id="loginsubmit"]')
submit.click()
for i in range(30):
time.sleep(1)
img = driver.find_element(By.XPATH, '/html/body/div[4]/div/div/div/div[1]/div[2]/div[1]/img')
img.screenshot("./img/{}.png".format(i))
time.sleep(0.5)
refresh = driver.find_element(By.XPATH, '/html/body/div[4]/div/div/div/div[1]/div[2]/div[2]')
refresh.click()
time.sleep(3)
driver.quit()（二）EasyDL机器学习（百度智能云）
（1）EasyDL零门槛AI开发平台-立即使用-物体检测（图像）
（2）创建模型:填写模型名称（自定义例如：myModel）+业务描述（自定义例如：Captcha Recognition)+完成创建
（3）数据总览中的创建数据集：填写数据集名称（自定义例如：myModelData）+完成
（4）数据总览中在上述数据集中：点击导入+无标注信息+导入方式：本地导入+上传图片（刚才代码抓取的所有图片）+确认并返回
（4）导入成功后在线标注：右上角添加标签（自定义例如：Target）后确定+标注每个图片的黑色区域即滑块移动位置
（5）标注好后训练模型：选择模型名称+添加数据集+公有云部署+开始训练+等待训练结束+查看模型效果+申请发布
（6）发布模型：公有云部署+提交申请
（三）调用EasyDLAPI接口识别验证码
（1）在上述模型中点击服务详情+查看API文档+点击EasyDL版控制台+公有云部署：应用列表+点击创建应用+输入应用名称+选择接口EasyDL+立即创建+获得此应用的API KEY和Secret Key
(2)再次来到API文档：导航栏中：模型发布-公有云部署-物体检测API调用文档中的复制代码创建新py文件修改代码如下:（通过main.png来测试）
import json
import base64
import requests
IMAGE_FILEPATH = "./img/main.png"
PARAMS = {"threshold": 0.3}
# 服务详情 中的 接口地址
MODEL_API_URL = "【您的API地址】"
# 调用 API 需要 ACCESS_TOKEN。若已有 ACCESS_TOKEN 则于下方填入该字符串
# 否则，留空 ACCESS_TOKEN，于下方填入 该模型部署的 API_KEY 以及 SECRET_KEY，会自动申请并显示新 ACCESS_TOKEN
# ACCESS_TOKEN = "【您的ACESS_TOKEN】"
ACCESS_TOKEN = None
API_KEY = "【您的API_KEY】"
SECRET_KEY = "【您的SECRET_KEY】"
print("1. 读取目标图片 '{}'".format(IMAGE_FILEPATH))
with open(IMAGE_FILEPATH, 'rb') as f:
base64_data = base64.b64encode(f.read())
base64_str = base64_data.decode('UTF8')
print("将 BASE64 编码后图片的字符串填入 PARAMS 的 'image' 字段")
PARAMS["image"] = base64_str
if not ACCESS_TOKEN:
print("2. ACCESS_TOKEN 为空，调用鉴权接口获取TOKEN")
auth_url = "https://aip.baidubce.com/oauth/2.0/token?grant_type=client_credentials" "&client_id={}&client_secret={}".format(API_KEY, SECRET_KEY)
auth_resp = requests.get(auth_url)
auth_resp_json = auth_resp.json()
ACCESS_TOKEN = auth_resp_json["access_token"]
print("新 ACCESS_TOKEN: {}".format(ACCESS_TOKEN))
else:
print("2. 使用已有 ACCESS_TOKEN")
print("3. 向模型接口 'MODEL_API_URL' 发送请求")
request_url = "{}?access_token={}".format(MODEL_API_URL, ACCESS_TOKEN)
response = requests.post(url=request_url, json=PARAMS)
response_json = response.json()
response_str = json.dumps(response_json, indent=4, ensure_ascii=False)
print("结果:
{}".format(response_str))云码平台一站式识别（实战：豆瓣网）
（一）获取验证码图片
from selenium import webdriver
from selenium.webdriver.chrome.service import Service
from selenium.webdriver.common.by import By
from selenium.webdriver import ActionChains
import time
import requests
service = Service(executable_path="../_resources/chromedriver")
driver = webdriver.Chrome(service=service)
driver.get("https://www.douban.com/")
time.sleep(2)
frame = driver.find_element(By.XPATH, '//*[@id="anony-reg-new"]/div/div[1]/iframe')
driver.switch_to.frame(frame)
time.sleep(0.5)
select = driver.find_element(By.XPATH, '/html/body/div[1]/div[1]/ul[1]/li[2]')
select.click()
username = driver.find_element(By.XPATH, '//*[@id="username"]')
ActionChains(driver).pause(0.5).click(username).send_keys("xxxxxxxxxx").parform()
password = driver.find_element(By.XPATH, '//*[@id="password"]')
ActionChains(driver).pause(0.5).click(password).send_keys("xxxxxxxxxx").parform()
submit = driver.find_element(By.XPATH, '/html/body/div[1]/div[2]/div[1]/div[5]/a')
submit.click()
time.sleep(2)
driver.switch_to.frame("tcaptcha_iframe_dy")
img = driver.find_element(By.XPATH, '//*[@id="slideBg"]')
img.screenshot("./img2/captcha.png")
time.sleep(3)
driver.quit()（二）验证码识别
云码-自动验证码识别平台：在线测试：点击单图滑块（截图）-上传刚获取的验证码图片-点击测试
以下为通过代码实现方法：登陆账号成功后点击个人中心-点击开发文档-点击python
修改文档代码如下（新建py文件名称为：ydm.py)：
注意根据个人情况修改token
import json
import time
import requests
import base64
class YdmVerify(object):
_custom_url = "http://api.jfbym.com/api/YmServer/customApi"
_token = ""
_headers = {
'Content-Type': 'application/json'
}
def common_verify(self, image, verify_type="60000"):
# 数英汉字类型
# 通用数英1-4位 10110
# 通用数英5-8位 10111
# 通用数英9~11位 10112
# 通用数英12位及以上 10113
# 通用数英1~6位plus 10103
# 定制-数英5位~qcs 9001
# 定制-纯数字4位 193
# 中文类型
# 通用中文字符1~2位 10114
# 通用中文字符 3~5位 10115
# 通用中文字符6~8位 10116
# 通用中文字符9位及以上 10117
# 定制-XX西游苦行中文字符 10107
# 计算类型
# 通用数字计算题 50100
# 通用中文计算题 50101
# 定制-计算题 cni 452
payload = {
"image": base64.b64encode(image).decode(),
"token": self._token,
"type": verify_type
}
print(payload)
resp = requests.post(self._custom_url, headers=self._headers, data=json.dumps(payload))
print(resp.text)
return resp.json()['data']['data']
def slide_verify(self, slide_image, background_image, verify_type="20101"):
# 滑块类型
# 通用双图滑块 20111
payload = {
"slide_image": base64.b64encode(slide_image).decode(),
"background_image": base64.b64encode(background_image).decode(),
"token": self._token,
"type": verify_type
}
resp = requests.post(self._custom_url, headers=self._headers, data=json.dumps(payload))
print(resp.text)
return resp.json()['data']['data']
def sin_slide_verify(self, image, verify_type="20110"):
# 通用单图滑块(截图) 20110
payload = {
"image": base64.b64encode(image).decode(),
"token": self._token,
"type": verify_type
}
resp = requests.post(self._custom_url, headers=self._headers, data=json.dumps(payload))
print(resp.text)
return resp.json()['data']['data']
def traffic_slide_verify(self, seed, data, href, verify_type="900010"):
# 定制-滑块协议slide_traffic 900010
payload = {
"seed": seed,
"data": data,
"href": href,
"token": self._token,
"type": verify_type
}
resp = requests.post(self._custom_url, headers=self._headers, data=json.dumps(payload))
print(resp.text)
return resp.json()['data']['data']
def click_verify(self, image, label_image=None, extra=None, verify_type="30100"):
# 通用任意点选1~4个坐标 30009
# 通用文字点选1(extra,点选文字逗号隔开,原图) 30100
# 定制-文字点选2(extra="click",原图) 30103
# 定制-单图文字点选 30102
# 定制-图标点选1(原图) 30104
# 定制-图标点选2(原图,extra="icon") 30105
# 定制-语序点选1(原图,extra="phrase") 30106
# 定制-语序点选2(原图) 30107
# 定制-空间推理点选1(原图,extra="请点击xxx") 30109
# 定制-空间推理点选1(原图,extra="请_点击_小尺寸绿色物体。") 30110
# 定制-tx空间点选(extra="请点击侧对着你的字母") 50009
# 定制-tt_空间点选 30101
# 定制-推理拼图1(原图,extra="交换2个图块") 30108
# 定制-xy4九宫格点选(原图,label_image,image) 30008
payload = {
"image": base64.b64encode(image).decode(),
# "label_image": base64.b64encode(label_image).decode(),
"token": self._token,
"type": verify_type
}
if extra:
payload['extra'] = extra
resp = requests.post(self._custom_url, headers=self._headers, data=json.dumps(payload))
print(resp.text)
return resp.json()['data']['data']
def rotate(self, out_ring_image, inner_circle_image):
# 定制-X度单图旋转 90007
# payload = {
# "image": base64.b64encode(image).decode(),
# "token": self._token,
# "type": "90007"
# }
# 定制-Tt双图旋转,2张图,内圈图,外圈图 90004
payload = {
"out_ring_image": base64.b64encode(out_ring_image).decode(),
"inner_circle_image": base64.b64encode(inner_circle_image).decode(),
"token": self._token,
"type": "90004"
}
resp = requests.post(self._custom_url, headers=self._headers, data=json.dumps(payload))
print(resp.text)
return resp.json()['data']['data']
def google_verify(self, googlekey, pageurl, invisible=1, data_s=""):
_headers = {
'Content-Type': 'application/json'
}
"""
第一步，创建验证码任务
:param
:return taskId : string 创建成功的任务ID
"""
url = "http://122.9.52.147/api/YmServer/funnelApi"
payload = json.dumps({
"token": self._token,
# "type": "40011", ## v3
"type": "40010", ## v2
"googlekey": googlekey,
"enterprise": 1, ## 是否为企业版
"pageurl": pageurl,
"invisible": invisible,
"data-s": data_s,
# 'action':"TEMPLATE" #V3必传
})
# 发送JSON格式的数据
result = requests.request("POST", url, headers=_headers, data=payload).json()
print(result)
# {'msg': '识别成功', 'code': 10000, 'data': {'code': 0, 'captchaId': '51436618130', 'recordId': '74892'}}
captcha_id = result.get('data').get("captchaId")
record_id = result.get('data').get("recordId")
times = 0
while times < 150:
try:
url = f"http://122.9.52.147/api/YmServer/funnelApiResult"
data = {
"token": self._token,
"captchaId": captcha_id,
"recordId": record_id
}
result = requests.post(url, headers=_headers, json=data).json()
print(result)
# {'msg': '结果准备中，请稍后再试', 'code': 10009, 'data': []}
if result['msg'] == "结果准备中，请稍后再试":
time.sleep(5)
times += 5
continue
if result['msg'] == '请求成功' and result['code'] == 10001:
print(result['data']['data'])
return result['data']['data']
# {'msg': '请求成功', 'code': 10001, 'data': {'data': '03AGdBq2611GTOgA2v9HUpMMEUE70p6dwOtYyHJQK4xhdKF0Y8ouSGsFZt647SpJvZ22qinYrm6MYBJGFQxMUIApFfSBN6WTGspk6DmFdQAoWxynObRGV7qNMQOjZ_m4w3_6iRu8SJ3vSUXH_HHuA7wXARJbKEpU4J4R921NfpKdahgeFD8rK1CFYAqLd5fz4l-8_VRmRE83dRSfkgyTN338evQ1doWKJRipZbk4ie-89Ud0KGdOsP4QzG3stRZgj2oaEoMDSAP62vxKGYqtDEqTcwtlgo-ot3rF5SmntaoKGwcKPo0NrekWA5gtj0vqKLU6lY2GcnSci_tgBzBwuH40uvyR1PFu02VK_E44mopJ7FOO4cUukNaLGqypU2YCA8QuaaebOIoCMU7RGqGs_41RYNCG1GSdthiwcwk2hHFbi-TXuICXSwh4Er5mgVW9A3t_9Ndp0eJcyr3HtuJrcA7BtlcgruuQxK5h4Ew4ert4KPH_aQGN9ww5VsUtbSManzUDnUOs7aEdvFk1DOOPmLys-aX20ZFN2CcQcZZSO-7HZpZZt3EDeWWE5S02HFDY8gl3_0xqIts8774Tr4GMVJaddG0NR6pcBFC11FqNcK2a18gM3gaKDy3_2ZMeSU4nj4NWwoAhPjQN2BS8JxX4kKVpX4rD959kc93vczVD3TYD6_4GJahGSpBvM7Y5_GGIdLL8imXde1R35mZnEcFYXQ40zcy3DdJFkk_gzGTVOEb1Q1IZpjMxzCxyGgwjgL9dtDIgst5H5CSZoerX_Lz-DmsBvYIYZdpbPLEMROx9MODImaEw8Cp6M8Xj7_foijiGE9hh-pzJSTlKl3HytiSUyJJ7r1BssrX5C_TFWxl0IXNg8azP8H-ZIOWwnYlMWCS1w9piHdoLg5zACiYIN3Txdlsvi61MuPmzJggJd1_dlyMdAlzb5_zdfweqj0_Ko1ODP378YT7sV7LECgRj5QJU6sF5nlf4m2g5sFypBw9GFAkEE-OaWGYxRJOy2ioU41ggAJIkcza2B_N5AL2KLROtm0-c2MxplM4ZzHxrUv9A24zlgzo3Pz4NONwU_gaOcDB7j1dZKXD8UaoIrZv0BTd8JeojYowm9Usdg7Rt4Fpo_vDLJdrEUfbxVlXieDD9Fr1fu72-d4AduT_J3n-rIhyX4gFav-KfP-qOxqOZsmjXZirsBxZs7042NYeirRYnLv35cxIAJARz03FJmeKViUivwC5mCWw64hjRad9XyyBOP2n8KFOrTXhPskC-WwEfksGtfLxi6VW76FHGvRdwHXzMwVfNqe3P5H_WZUc-vxeTAsTnqZz3WA97lM4MLrX0nTZYgXxCEiS6raSOiEMqcx_Nv7Zxre-abj4LZRbFpH8nx1SEiaOV2Dm-a1iPFEmCs0L4kDtt6VImSVIQaTOAd3KFSo7W_XTvRPsQJOtblrcKyuagztX_Yr0lT0YqN9I9MZAARo7M5OfwSLJW16rdmp4NuRefEvNPNHO2cVh1Xha1qNGuF_QDvWFFmWG0Y6IbRqLmF-Dv8BY4TWyOeVnADJftGQw2QSr8RmbCHryA'}}
except Exception as e:
print(e)
continue
def fun_captcha_verify(self, publickey, pageurl, verify_type="40007"):
# 定制类接口-Hcaptcha 40007
payload = {
"publickey": publickey,
"pageurl": pageurl,
"token": self._token,
"type": verify_type
}
resp = requests.post(self._custom_url, headers=self._headers, data=json.dumps(payload))
print(resp.text)
return resp.json()['data']['data']
def hcaptcha_verify(self):
# 定制类接口-Hcaptcha
_headers = {
'Content-Type': 'application/json'
}
_custom_url = "http://api.jfbym.com/api/YmServer/funnelApi"
payload = {
"sitekey": "",
"pageurl": "",
"token": self._token,
"type": '50013'
}
result = requests.post(_custom_url, headers=_headers, data=json.dumps(payload)).json()
print(result)
captcha_id = result.get('data').get("captchaId")
record_id = result.get('data').get("recordId")
times = 0
while times < 150:
try:
url = f"http://api.jfbym.com/api/YmServer/funnelApiResult"
data = {
"token": self._token,
"captchaId": captcha_id,
"recordId": record_id
}
result = requests.post(url, headers=_headers, json=data).json()
print(result)
# {'msg': '结果准备中，请稍后再试', 'code': 10009, 'data': []}
if result['msg'] == "结果准备中，请稍后再试":
time.sleep(5)
times += 5
continue
if result['msg'] == '请求成功' and result['code'] == 10001:
print(result['data']['data'])
return result['data']['data']
# {'msg': '请求成功', 'code': 10001, 'data': {'data': '03AGdBq2611GTOgA2v9HUpMMEUE70p6dwOtYyHJQK4xhdKF0Y8ouSGsFZt647SpJvZ22qinYrm6MYBJGFQxMUIApFfSBN6WTGspk6DmFdQAoWxynObRGV7qNMQOjZ_m4w3_6iRu8SJ3vSUXH_HHuA7wXARJbKEpU4J4R921NfpKdahgeFD8rK1CFYAqLd5fz4l-8_VRmRE83dRSfkgyTN338evQ1doWKJRipZbk4ie-89Ud0KGdOsP4QzG3stRZgj2oaEoMDSAP62vxKGYqtDEqTcwtlgo-ot3rF5SmntaoKGwcKPo0NrekWA5gtj0vqKLU6lY2GcnSci_tgBzBwuH40uvyR1PFu02VK_E44mopJ7FOO4cUukNaLGqypU2YCA8QuaaebOIoCMU7RGqGs_41RYNCG1GSdthiwcwk2hHFbi-TXuICXSwh4Er5mgVW9A3t_9Ndp0eJcyr3HtuJrcA7BtlcgruuQxK5h4Ew4ert4KPH_aQGN9ww5VsUtbSManzUDnUOs7aEdvFk1DOOPmLys-aX20ZFN2CcQcZZSO-7HZpZZt3EDeWWE5S02HFDY8gl3_0xqIts8774Tr4GMVJaddG0NR6pcBFC11FqNcK2a18gM3gaKDy3_2ZMeSU4nj4NWwoAhPjQN2BS8JxX4kKVpX4rD959kc93vczVD3TYD6_4GJahGSpBvM7Y5_GGIdLL8imXde1R35mZnEcFYXQ40zcy3DdJFkk_gzGTVOEb1Q1IZpjMxzCxyGgwjgL9dtDIgst5H5CSZoerX_Lz-DmsBvYIYZdpbPLEMROx9MODImaEw8Cp6M8Xj7_foijiGE9hh-pzJSTlKl3HytiSUyJJ7r1BssrX5C_TFWxl0IXNg8azP8H-ZIOWwnYlMWCS1w9piHdoLg5zACiYIN3Txdlsvi61MuPmzJggJd1_dlyMdAlzb5_zdfweqj0_Ko1ODP378YT7sV7LECgRj5QJU6sF5nlf4m2g5sFypBw9GFAkEE-OaWGYxRJOy2ioU41ggAJIkcza2B_N5AL2KLROtm0-c2MxplM4ZzHxrUv9A24zlgzo3Pz4NONwU_gaOcDB7j1dZKXD8UaoIrZv0BTd8JeojYowm9Usdg7Rt4Fpo_vDLJdrEUfbxVlXieDD9Fr1fu72-d4AduT_J3n-rIhyX4gFav-KfP-qOxqOZsmjXZirsBxZs7042NYeirRYnLv35cxIAJARz03FJmeKViUivwC5mCWw64hjRad9XyyBOP2n8KFOrTXhPskC-WwEfksGtfLxi6VW76FHGvRdwHXzMwVfNqe3P5H_WZUc-vxeTAsTnqZz3WA97lM4MLrX0nTZYgXxCEiS6raSOiEMqcx_Nv7Zxre-abj4LZRbFpH8nx1SEiaOV2Dm-a1iPFEmCs0L4kDtt6VImSVIQaTOAd3KFSo7W_XTvRPsQJOtblrcKyuagztX_Yr0lT0YqN9I9MZAARo7M5OfwSLJW16rdmp4NuRefEvNPNHO2cVh1Xha1qNGuF_QDvWFFmWG0Y6IbRqLmF-Dv8BY4TWyOeVnADJftGQw2QSr8RmbCHryA'}}
except Exception as e:
print(e)
continue
新建py文件：
from ydm import YdmVerify
Y = YdmVerify()
with open('./img2/captcha.png','rb') as f:
s = f.read()
Y.sin_slide_verify(image=s)（三）动态获取验证码
根据一和二结合代码，修改py文件
from selenium import webdriver
from selenium.webdriver.chrome.service import Service
from selenium.webdriver.common.by import By
from selenium.webdriver import ActionChains
import time
import requests
from ydm import YdmVerify
service = Service(executable_path="../_resources/chromedriver")
driver = webdriver.Chrome(service=service)
driver.get("https://www.douban.com/")
time.sleep(2)
frame = driver.find_element(By.XPATH, '//*[@id="anony-reg-new"]/div/div[1]/iframe')
driver.switch_to.frame(frame)
time.sleep(0.5)
select = driver.find_element(By.XPATH, '/html/body/div[1]/div[1]/ul[1]/li[2]')
select.click()
username = driver.find_element(By.XPATH, '//*[@id="username"]')
ActionChains(driver).pause(0.5).click(username).send_keys("xxxxxxxxxx").parform()
password = driver.find_element(By.XPATH, '//*[@id="password"]')
ActionChains(driver).pause(0.5).click(password).send_keys("xxxxxxxxxx").parform()
submit = driver.find_element(By.XPATH, '/html/body/div[1]/div[2]/div[1]/div[5]/a')
submit.click()
time.sleep(2)
driver.switch_to.frame("tcaptcha_iframe_dy")
img = driver.find_element(By.XPATH, '//*[@id="slideBg"]')
img.screenshot("./img2/captcha.png")
time.sleep(1)
Y = YdmVerify()
with open('./img2/captcha.png','rb') as f:
s = f.read()
Y.sin_slide_verify(image=s)
time.sleep(3)
driver.quit()