python调用人脸识别平台接口并解析json

将本地图片转化为base64给face++接口,得到返回的json并解析

1. 安装requests库
   为什么选择这个库 ? Requests它会比urllib更加方便，可以节约我们大量的工作。requests是python实现的最简单易用的HTTP库，建议爬虫使用requests库。默认安装好python之后，是没有安装requests模块的，需要单独通过pip安装(请先安装pip)
   pip insatll request

2. 编写py代码

# coding=utf-8
# face++ 人脸识别detect部分
import requests
import base64
import json

def file_base64(file_name):
    with open(file_name, 'rb') as fin:
        file_data = fin.read()
        base64_data = base64.b64encode(file_data)
    return base64_data

def post_pic():
    url = 'https://api-cn.faceplusplus.com/facepp/v3/detect'
    image_1 = file_base64('/home/dashuaibi/图片/1.jpg') 

    postData = {
        'api_key': 'mTxxxxxxxxxxxxxxxxwuG',
        'api_secret': 'uo3xxxxxxxxxxxxxxxxxxxxxxxxeb5',
        'image_base64': image_1,
        'return_attributes' : 'emotion'
    }

    back = requests.post(url, data = postData)
    jres = json.loads(back.text)
    print jres['faces'][0]['attributes']['emotion']

if __name__=="__main__":
    post_pic()

3. 运行后发现打印的是
   {u'neutral': 60.45, u'sadness': 0.963, u'disgust': 0.104, u'anger': 3.324, u'surprise': 0.795, u'fear': 0.104, u'happiness': 34.26}
   在loads后变成python变量时，元素都变成了单引号，并且字符串前加多了个u
   json.loads()会在key前加上Unicode识别字符
   但是只是提示作用而已 并不影响数据

4. 编写业务逻辑

# coding=utf-8
# face++ 人脸识别detect部分
import requests
import base64
import json

def file_base64(file_name):
    with open(file_name, 'rb') as fin:
        file_data = fin.read()
        base64_data = base64.b64encode(file_data)
    return base64_data

def post_pic():
    url = 'https://api-cn.faceplusplus.com/facepp/v3/detect'
    image_1 = file_base64('/home/dashuaibi/图片/1.jpg') 

    postData = {
        'api_key': 'mTtxxxxxxxxxxxxxwuG',
        'api_secret': 'uo3xxxxxxxxxxxxxxJeb5',
        'image_base64': image_1,
        'return_attributes' : 'emotion'
    }

    back = requests.post(url, data = postData)
    jres = json.loads(back.text)
    #emotion各可能性
    info = jres['faces'][0]['attributes']['emotion']
    anger = info['anger']
    disgust = info['disgust']
    fear = info['fear']
    happiness = info['happiness']
    neutral = info['neutral']
    sadness = info['sadness']
    surprise = info['surprise']
    emotion = [anger,disgust,fear,happiness,neutral,sadness,surprise]
    emotion_name = ['愤怒','厌恶','恐惧','高兴','平静','伤心','惊讶']
    print '您现在的表情是'+emotion_name[emotion.index(max(emotion))]

if __name__=="__main__":
    post_pic()

5. 输出结果 :

6. 总结:

• python的库超级强大!!!!!!
• pip超级好用!!!!!
• 因为是linux的gedit手打,相当于记事本了,就懒得每一行写注释了嘿嘿