Skip to content

Camera Connector

The Camera Connector page is the app's first page. It says whether the app is running, connects cameras and sets when each camera takes a picture.

Camera Connector: Is it running?, the connected camera, and its live preview

Is it running?

The card at the top answers in one line each:

Line Says
Cameras Each camera, its trigger mode and pictures per second
Running Each detector and its camera, for example Color Detection: orange, red on camera 1
Licence Whether the Runtime licence is held
Data Layer insiso/visual-detector/, whether it is ready, and the pictures evaluated

A ✘ comes with the reason and what to do.

Connect a camera

  1. Under Connect, choose the Connection type: GenICam / GigE Vision, USB / UVC camera, RTSP stream or HTTP stream.
  2. For a GigE Vision camera, press Scan for cameras and pick it. For a stream, type its address.
  3. Press Connect.

The camera appears under Connected cameras with its state, resolution and frame rate. A color detector starts on it at once, and the loaded model runs on it too. Connecting a second camera adds it beside the first. Each connected camera needs its own Camera Connector licence.

After a reboot or a camera power cycle the app reconnects the camera by itself, with its trigger mode.

When the camera takes a picture

Set Takes a picture for each camera:

Setting Use
continuously (free-run) The camera streams; every picture is evaluated. For a moving belt and for commissioning
when triggered (software) One picture each time the PLC writes trigger, or you press Trigger once
on its trigger line (hardware) One picture per pulse on the camera's trigger input. GenICam cameras only

A software trigger fires every connected camera at the same moment. The answer appears on the Data Layer; see Data Layer.

Preview

Preview shows one camera's live picture. Use the camera buttons to switch. Only one preview runs at a time, so the page stays fast.

Saved pictures

Pictures are not stored unless you ask for it.

Saved pictures: take one now, or save on detection, to the device or a USB drive

  • Take a picture now saves one picture.
  • Save automatically when something is detected saves a picture when a detector finds something, at most one every few seconds (At most one every (s)). Only these classes limits it to some classes.
  • Save every picture (free run), to a drive only keeps every picture, on a USB drive.
  • Save chooses the detection overlay, the raw camera picture (for retraining), or both.
  • Keep at most and Disk budget (MB) cap the buffer. The oldest pictures go first.
  • Download all (.zip) downloads every saved picture with its detections. See Captures for retraining.

HMI stream link gives a video-only page of a camera, for a ctrlX HMI web-view widget. Opening it never slows the detection.