HTTP Integration

Function path: SVE navigation bar → Result Integration → HTTP

---

Overview

The smart camera reports AI inference results to an HTTP server via REST API (POST). The camera sends result messages in JSON format (containing image links originImg / finalImg / maskImg), and the server receives and archives them as agreed.

Integration Requirements

Item	Requirement
HTTP Server	Accessible by the camera (same network segment), standard HTTP service deployed, with an endpoint to receive reports
Request Method	POST (camera-fixed)
Content-Type	application/json (camera-fixed)
Request Path	http://<ServerIP>:<Port>/<EndpointPath>
Data Content	JSON: {"appId","appVersion","cameraId","timestamp","originImg","maskImg","finalImg","additional"}

Basic Configuration on SVE

1. Navigate to Result Integration → HTTP;
2. Turn on the HTTP Integration switch;
3. Fill in REST API: <ServerIP>:<Port>/<EndpointPath>;

Note: The "REST API" input field in the SVE interface already includes the http:// prefix. You only need to enter the IP, port, and endpoint path — no need to add http:// again.

4. Request Method = POST (fixed, cannot be changed); Content-Type = application/json (fixed, cannot be changed);
5. Click Save.

Basic Verification

Page Testing (Test Send)

1. After configuration, click Test Send;
2. Observe the "Debug Request" area on the right — an HTTP 200 response indicates the link is normal.

Real Task Verification

1. In Task Management → AI Apps, trigger "Run Preview" for the application to let the camera complete one inference;
2. Confirm on the HTTP server/receiving end that the camera's report request is received: content is complete JSON (containing appId / timestamp / originImg and other image links) → Integration successful.

Common Issues

Symptom	Troubleshooting
Camera report failed	Check if port/IP is reachable; if the server endpoint is running (listening port)
Receiving end reports format error	Confirm the receiving end parses as application/json; request path matches the server endpoint
No report received	Confirm HTTP Integration switch is on and saved; inference has been triggered; REST API correctly filled with IP/port/endpoint path