To facilitate data transfer between the camera connector and vision-driven applications, the currently supported communication channels are ZeroMQ (ZMQ) and MQTT. For enhanced integration, such as with Siemens Industrial Edge, supplementary interfaces like REST API could also be considered.
The connector application must possess two bind ZMQ ports and/or MQTT configurable connection parameters. These connections will be utilized for setting up the application, configuring the camera, defining image streaming options, and requesting images via software triggers. Images must adhere to the Vision Connector Image Format.
All used ports and servers must be configurable in the application configuration file (for Siemens Industrial Edge applications) and environment variables. The ZMQ ports (defaults: 5555 to REQ/REP and 5556 to PUB/SUB) must be documented in the application documentation.
In the case of MQTT connections, the server, port, user, password, client id and root topic should be configurable within the application.
The connector should be designed to support one or multiple cameras. This support must be thoroughly detailed in the application documentation and should include a dedicated topic structure for each camera.
The scope of this document is to define the interface protocols between vision-driven applications and camera connectivity applications. Communication specifics, camera protocols, camera setup, and application features are beyond the scope of this document.
Below is a schematic representation of the defined scope: