04-10-2017 03:37 AM
Hello I would like the error detection on technical surface (texture extraction image processing). I will use Gabor filters to analyze 8 bit images. I do not know how to create in LabView code, a VI for 2D images Gabor filter. Input data are 8 bit images. Many thanks for your help
Solved! Go to Solution.
04-10-2017 03:45 AM
Hello I have found this Vi. But my input data are 2d images with 8bit. Now how can i customize this vi. (See appendix Gabor.JPG)
Thank you very much.
04-10-2017 04:13 AM
The enclosed VI is unfortunately not yet adapted for 2D images.
I have only printed out on provided on. Your help is still needed.
04-10-2017 03:24 PM
Hi Bari,
That is interesting issue you have.
I can see the Gabor filter in the advance signal processing. But I don't see it in the library.
But to my understanding Gabor kernel can be design in another package like matlab.
Then you can do convolution with this Kernal.
04-11-2017 02:06 AM
Hello,
Yes they are right with Matlab. But it would be better if we would work together for Labview too.
Especially Gabor Filter is of great importance for image processing.
Thank you very much.
04-12-2017 02:08 PM
Hi,
I notice that NI has new package to run Open CV.
Open CV has Gabor filter bank.
http://www.ni.com/white-paper/53072/en/
04-13-2017 05:12 AM
Hello shachaf,
Thanks for the note with Open CV.
You have experience with Gabor Filler and transformation for image processing.
04-13-2017 01:20 PM
I wrote couple of applications that use texture analysis.
I usually use statistical feature descriptors calculate base on localize histogram and not Gabor.
LV has texture calculation that I think is based on statistical feature calculation.
But since you can't open the source code in LV I prefer to write my own functions.
I can give you the code that calculate the statistical feature descriptors.
I would be interested if you are planning to write the Gabor filter bank.
There is source code in Matlab avalible:
http://www.mathworks.com/matlabcentral/fileexchange/44630-gabor-feature-extraction