模型的预测结果格式是base64图像，输出格式是：

{
    "results": [
        "iVBORw0KGgoAAAANSUhEUgAAAgAAAAIACAIAAAB7GkOtAAEAAEl..."
    ]
}

部署AI应用（在线服务）完成后，可通过页面测试服务，

也可通过API调用方式访问在线服务，可参考访问在线服务简介。

示例1：使用Python语言发送预测请求（AK/SK认证）

import requests
from apig_sdk import signer

# Config url, ak, sk and file path.
url = "在线服务调用地址"
ak = "AK值"
sk = "SK值"
file_path = "预测文件的本地路径"

# Create request, set method, url, headers and body.
method = 'POST'
headers = {"x-sdk-content-sha256": "UNSIGNED-PAYLOAD"}
request = signer.HttpRequest(method, url, headers)

# Create sign, set the AK/SK to sign and authenticate the request.
sig = signer.Signer()
sig.Key = ak
sig.Secret = sk
sig.Sign(request)

# Send request
files = {'images': open(file_path, 'rb')}
resp = requests.request(request.method, request.scheme + "://" + request.host + request.uri, headers=request.headers, files=files)

# Print result
# print(resp.status_code)
# print(resp.text)

# base64图片显示

import base64
from io import BytesIO
from PIL import Image
import matplotlib.pyplot as plt

result = eval(resp.text)
base64imgs = result["results"]
for i in base64imgs:
    imgdata = base64.b64decode(i)
    img = Image.open(BytesIO(imgdata))
    plt.imshow(img)
    plt.show()