> I am new to Computer Vision, creating a image search application using Zernike Moments as feature
> I would like to know below
> 1) Is mahotas support MKL or GPU if yes how to enable it or where can i get these binaries
Only through the operations that it uses numpy for (which is a minority of them).
> 2) What is the relation of pixel intensity with ZM
> please help me as i am stuck and the project is at halt
>
> Observation
> Same image with different intensity produces largely different ZM
Yes, that is to be expected. The Zernike moments are a deconstruction of the intensity image, not just the shape. Thus, different intensitities will produce different values. In some contexts, binarizing (by thresholding) is a good preprocessing step.
Cheers!
Luis
>
>
>
> --
> You received this message because you are subscribed to the Google Groups "pythonvision" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to
pythonvision...@googlegroups.com.
> For more options, visit
https://groups.google.com/d/optout.