Feature based geometric match stability issue

[ChrisStrykesAgain]

desalando,

Are you experiencing a similar issue? I don't see any progress on the CAR but would love to assist you if you are experiencing similar issues, and can re-escalate to R&D if it proves neccessary.

[desalando]

Hello Christopher,

Yes, i think i have a similar problem. There is a topic that we discouse about it.

http://forums.ni.com/t5/Machine-Vision/Instability-in-pattern-matching/td-p/2876274/page/3

AndyN wrote:

(1)"I am trying to use Feature and  Edge based pattern matching.  With both methods I get the same problem of lack  of consistency in the match results.

In one image the score is ie 990 and in the next frame there is no recognition even if the threshold of score is set to zero.

The difference between the images is marginal. They were taken 16.7 ms apart with close to a stationary object. "

(2) "What it seems is wrong here is the fact that the method/math/algorithm does not return the similarity correctly. It says C is similar to B at zero level.

If this is true, than there is no point in using that method. It is a binary method and the correlation factor means nothing.

It does not matter what feature I use and how good the image is. Your hope is as good of a measure of robustness.

You come up with a set of feature on two almost identical images, they will fail on the third, three, it will fail on the fourth, and so on.

[ChrisStrykesAgain]

Desalando, I glanced through the thread you linked and it seems very similar to this issue. What differences, if any, are there between the issues discussed in the two threads?

[desalando]

Hello Christopher,

Sorry to bother you. I'm saying the same thing. 

Joel writes on ‎01-18-2013 that "I reported this issue to our R&D Team under CAR #384585, the title is "Feature based geometric match instability" and it is currently under investigation."

Actually my question was, if there is any solution for CAR #384585 and generally "instability".

I thing if there was a CAR and "investigation", after one and a half years later, there must be a solution now for this issue.  

However, today I have same problems with Edge based geometric match. I have latest versions (NI Vision Development Module 2013 SP1, NI Vision Acquisition Software February 2014)

[ChrisStrykesAgain]

My apologies. I don't know if I was looking at the wrong CAR or am just blind, but the issue was accepted as a valid CAR under that number and the target version for the fix is listed as 2013 SP1.

I don't see it explicitly listed in the release notes but I will follow up to see what the status of this issue is.

[AndyN]

It would be nice to know it was fixed and having some sort of trail of actions.

I can't believe NI released the module with such a bug.

I see this very dissapointing pattern of users being utilized as free labor testers. What is even worse they have to push the support team to the wall to admit of having an issue and taking actions. Denial is not a proper attitude.

Cheers,