Why and when continous triggering?

The continuous triggering in Scorpion was designed to make it possible to captures 10 – 200 images/second. The initial and simple image aquisition model defined by this pattern:

  1. Issue the Grab command
  2. Wait for all ImageComplete events
  3. Issue the Inspect to process the aquired images
  4. Use the AfterInspection event to do IO
  5. Goto 1 or Stop the inspection cycle

Read the details explanation about the event-driven model in Scorpion – http://scorpion.tordivel.no/help/UsersGuide/Service/Actions/Default.htm

The simple model has a drawback when the camera is free-running or the system is hardware triggered if you want to run fast. The Grab command takes time and can be eliminated. This changes the basic models from:

  1. Issue the Grab command
  2. Wait for all ImageComplete events
  3. Goto 1

to the continous model

  1. Wait for all ImageComplete events
  2. Goto 1

The difference is the Grab command is implicit and not explicit in the continous model. The continous model is supported by most Scorpion Camera Drivers.

The implicit Grab command saves milliseconds which is important when running fast 20-100 images / second with a cycle time of 10 – 50 ms. It is more important the faster it goes.

The drawback with the continuous model is that Scorpion Vision might the pushed to hard – too many images might flow into the system and what we call “cpu-starvation” might occur. This may happen if images are pushed to hard into Scorpion or maybe the is noise on the hardware trigger input.

Important Rule of thumb:  The basic image capture model is the best if you do not need the speed of the continuous model.

%d bloggers like this: