10-15-2014 04:29 AM - edited 10-15-2014 04:33 AM
Hi All,
I am exploring the LabVIEW machine learning toolkit available at https://decibel.ni.com/content/docs/DOC-19328.
I want to try the unsupervised k-means clustering on an image to segment it.
I have a color image of an apple. I want to be able to segment the defects on the surface based on the difference in the color using K means. There is a brown patch on the surface of the apple, I want to be able to segment it out using k means.
However I am unable to figure out how to proceed. I checked the examples provided with the toolkit however, they dont use images.
Has anyone worked on this toolkit? Any guidance can be very helpful.
Thanks.
Regards.
Aveo
10-15-2014 08:26 AM - edited 10-15-2014 08:27 AM
Hello,
looking at the image you've posted, it is going to be difficult to segment the brown patch reliably. You should first make sure that the camera is focused properly and perhaps you could somehow use additional lighting to your advantage.
But anyway, see an example of image segmentation based on K-means clustering (using the Labview machine learning toolkit) here:
I have just posted this in response to your post. I hope you don't mind and that it will be of some use to you (and others).
Best regards,
K
10-16-2014 12:03 AM
Hi Klemen,
Thank you for the blog post. It has helped a lot to understand the toolkit for k means clustering.
Its awesome to see you contribute for the greater good of everyone. Please continue the same.
Regards
Aveo