python小程序完成单人脸识别和多人脸识别（超有意思）

1、单人脸识别

# 导入需要用的模块
from aip import AipFace

# 准备AppID、API Key、Secret Key
APP_ID = '18101974'
API_KEY = 'HG5NFSFpfBiqqmBYVBpXQFKs'
SECRET_KEY = 'ytdOFrcBxZvvGlMKGtmkmpBQ6arbjupx'

# 调用人脸检测
client = AipFace(APP_ID, API_KEY, SECRET_KEY)

# 图片的类型，三种可选BASE64、URL、FACE_TOKEN
image_type = 'URL'
# 图片的url
iamge = 'https://timgsa.baidu.com/timg?image&quality=80&size=b9999_10000&sec=1577292322895&di=f7fa123b95016762f1f0c3d2999328c4&imgtype=0&src=http%3A%2F%2Fww1.sinaimg.cn%2Flarge%2Fc7b64b24jw1f7k7i3aap4j20qo1407ck.jpg'

# 识别人脸
face_data = client.detect(iamge, image_type)
# 导入需要用的模块
from aip import AipFace

# 准备AppID、API Key、Secret Key
APP_ID = '18101974'
API_KEY = 'HG5NFSFpfBiqqmBYVBpXQFKs'
SECRET_KEY = 'ytdOFrcBxZvvGlMKGtmkmpBQ6arbjupx'

# 调用人脸检测
client = AipFace(APP_ID, API_KEY, SECRET_KEY)

# 图片的类型，三种可选BASE64、URL、FACE_TOKEN
image_type = 'URL'
# 图片的url
iamge = 'https://timgsa.baidu.com/timg?image&quality=80&size=b9999_10000&sec=1577292322895&di=f7fa123b95016762f1f0c3d2999328c4&imgtype=0&src=http%3A%2F%2Fww1.sinaimg.cn%2Flarge%2Fc7b64b24jw1f7k7i3aap4j20qo1407ck.jpg'

# 识别人脸
face_data = client.detect(iamge, image_type)
# 从Pillow中导入Image模块
from PIL import Image
import urllib

# 将image保存到本地文件
resp = urllib.request.urlopen(iamge)
f = open('im.jpg', 'wb')
f.write(resp.read())

# 使用Image模块，打开刚刚保存的图像
im = Image.open('im.jpg')

# 获取face_data中人脸的位置
left = face_data['result']['face_list'][0]['location']['left']
top = face_data['result']['face_list'][0]['location']['top']
width = face_data['result']['face_list'][0]['location']['width']
height = face_data['result']['face_list'][0]['location']['height']

# 截取人脸
face_im = im.crop((left, top, left+width, top+height))
face_im.save('face_im.jpg')

# pip install --user pillow -i https://pypi.tuna.tsinghua.edu.cn/simple

人脸识别前图片：

人脸识别后照片： 

2、多人脸识别

# 导入需要用的模块
from aip import AipFace

# 准备AppID、API Key、Secret Key
APP_ID = '18101974'
API_KEY = 'HG5NFSFpfBiqqmBYVBpXQFKs'
SECRET_KEY = 'ytdOFrcBxZvvGlMKGtmkmpBQ6arbjupx'

# 调用人脸检测
client = AipFace(APP_ID, API_KEY, SECRET_KEY)

# 图片的类型，三种可选BASE64、URL、FACE_TOKEN
image_type = 'URL'
# 图片的url
image = 'https://timgsa.baidu.com/timg?image&quality=80&size=b9999_10000&sec=1577296093031&di=f88c884ea4440935055c704fd20c0549&imgtype=0&src=http%3A%2F%2F5b0988e595225.cdn.sohucs.com%2Fimages%2F20180209%2F3a8e9aca0b8c481298246a26451bf242.jpeg'

# 准备参数
options = {}
options['max_face_num'] = '6'

# 识别人脸，比之前多了个options，options为字典类型
face_data = client.detect(image, image_type, options)

# 从Pillow中导入Image模块
from PIL import Image
import urllib

# 将image保存到本地文件
resp = urllib.request.urlopen(image)
f = open('im.jpg', 'wb')
f.write(resp.read())

# 使用Image模块，打开刚刚保存的图像
im = Image.open('im.jpg')

# 用于区分图片
n = 0
# 遍历face_list
for i in face_data['result']['face_list']:
    n += 1
    left = i['location']['left']
    top = i['location']['top']
    width = i['location']['width']
    height = i['location']['height']

    # 截图并保存
    face_im = im.crop((left, top, left + width, top + height))
    face_im.save('face_im' + str(n) + '.jpg')

识别图片如下： 

 人脸识别结果如下：