How can we apply fuzzy logic to pattern recognition?

[pinkman]

Good day, we are a little bit confused about fuzzy logic and where it can be best applied in a pattern recognition program, so we are wondering where it can be best applied since our project is comparing two PCB(Printed Circuit Board) images using Imaq functions and etc.

Our program compares the two PCB images, one taken from an actual photo by a webcam and another in .jpeg file converted from Eagle .brd file. The two images is then thresholded then compated by getting its percent difference and also the number of open and short circuited coppers will be classified.

Is it possible to apply fuzzy logic prior to the thresholding function?

Thank you.

[Moorethoughts]

Pinkman,

We have a few resources that you might find handy for pattern recognition that do not utilize the fuzzy logic toolkit.  These patterns are based off of a set template, and they can be based off of geometric patterns or color patterns.  Please reference the "Pattern Matching" section of the following document for more information:

http://www.ni.com/white-paper/3470/en

Beyond that, if you are still looking to specifically utilize functions similar to those offered by the fuzzy logic toolkit, you could look at the IMAQ Classification palette, which allow you to identify unknown objects by comparing significant features to a set of features that conceptually represent classes of known objects.  This palette can be found in Vision and Motion » Image Processing » Texture » Classification.  As far as actually utilizing the fuzzy logic toolkit is involved, I do not believe that it is possible to use those functions for image processing.  Hopefully this gives you a bit of information on techniques you can use from here on out.  Please let me know if this answers your questions.

Regards,

Keith M.

Applications Engineering