Detecting pinhole

[loverttsl]

Hi,

I am doing a project using Vision Assistant and LabView8.6.

The purpose is detecting pinhole in both white and black background.

Initially, my idea was use "Quantify" the ROI.

But, it is found that this approach is too sensitive as even 1 pixel will FAIL my test.

And also loading and unloading the unit will cause the position of unit slightly differ from each other.

Is there any better way to do it??

Attached is the sample image i have.

Thanks in advance.

[muks]

Use a grayscale erode. Check the attached vi

[loverttsl]

Hi Muks,

Firstly thanks for your fast response.

By using the vi you provided, i can "see" the defect clearly.

But i have another concern is "How to make the defect a measurable unit"?

And attached also the real image i grabbed from camera with controlled ambient light.

Another request is can you upload as well how to achieve it by the mean of Vision Assistant 2009?

Thanks in advance.

[muks]

---

By using the vi you provided, i can "see" the defect clearly.

But i have another concern is "How to make the defect a measurable unit"?

---

Play around with particle filter and particle analysis.

[muks]

Where is the pin hole in the attached picture? One easy way if you are looking for pin hole in the inner area is to mask it in such a way that the pin hole alone remains when you do "Remove border objects".

[loverttsl]

Hi Muks,

"One easy way if you are looking for pin hole in the inner area is to mask it in such a way that the pin hole alone remains when you do "Remove border objects".

That is correct if i already know where the pinhole is.

But my problem is i need to find out whether there is any pinhole occurred or not.

Is it possible that you upload a screen capture of how you do it in Vision Assistant?

Is "Particle Analysis" and "Particle filtering" able to capture printing error (please see attached example) and capture pinhole that exactly attached to border(please see attached example)?

Thanks and Regards.

[muks]

Please confirm whether you are looking for only defects in the inner "Ring" (actually the inner arc) if we can call it that way?

[loverttsl]

If your "Inner Ring" is refer to the Area inside the Red Circle (please refer to attachment), then my answer is YES.

By the way, Muks,are you using MSN?

If we can discuss over MSN i think it will be more efficiency.

[muks]

Great that gives me claritty. I will work on that and get back. Alas i am not on msn..

[loverttsl]

Hi Muks, Any updates? hehehe......