Vision Builder 2010 SP1 image logging for object detection

[MariusSica]

Hi there!

I'm using Vision Builder 2011 SP1 to test if an segment display lights up correctly.

The test sequence, uses Vision Assistant and Detect Objects steps.

At the end of the sequence I've added an Image Logging step in order to save the images, but only when the inspection status fails.

The issue I'm having, is that I can only save pictures for the Vision Assistant steps, so if the display fails at any of the Detect Objects steps, I get nothing and I have to manually retest the display.

How can I obtain image logs if an Detect Objects step fails?

Thank you in advance, and have a great day!

[ChristopheC]

Use the Set Inspection Status step prior to calling the image logging step. This will update the Inspection Status variable that can be used as a condition to log the images or not in the Image logging step (Log Image Only When Inspection Status Fails, which is selected in your screenshot).

For this simple case, that would be the recommended way of doing it.

If you want to do more things if the inspection fails, I encourage you to look at how the Vision Builder AI state diagram works.

You can create pass and fail states with transitions based on a measurement made in a previous state.

Take a look at the example C:\Users\Public\Documents\National Instruments\Vision Builder AI 2015\Examples\Tutorial 5 - State Diagram Branching.vbai

And the corresponding instructions in the tutorial:

C:\Users\Public\Documents\National Instruments\Vision Builder AI 2015\Documentation\VBAI_Tutorial.pdf

(Chapter 6: Branching and Decision Making).

Hope this helps.

Christophe

[MariusSica]

Hi Christophe, thanks for your feedback.

I can get the "Image Logging" step to log images if any of the previous steps are FAIL, so no problems there.

My issue, is that I cannot set it up to log images from "Detect Objects" steps (these are the only steps in my sequence, that actually highlight the problem of the display).

I can only log images from the "Image/Simulate acquisitions" and/or "Vision Assistant", and this is not helpful, because they show the entire display without any highlights, leaving me to manually search for the needle in the haystack.

Thanks!

[ChristopheC]

Only some steps in Vision Builder AI create or modify the image and make a copy of them, in order for you to retrieve it later with a Select Image step, or access them in the image logging step. As the Detect object does not modify the image, the image at this specific point is not available for logging.

If you want to log the image that is processed by the Detect Object step, you will need to log the previous step that modifies - and makes an internal copy of the image: in your case, the Vision Assistant step that is the step just before the Detect Object step.

It doesn't look like we're documenting which steps are image producers, but they are essentially all the acquisition steps, and steps that modify the content of the image, such as the Vision Assistant step and the Threshold Image steps.

Hope this answers your question.

Best regards,

Christophe