Re: Filtering Signals according to an Amplitude Range

[Juan Pablo Carbajal]

On Fri, Dec 7, 2012 at 3:01 PM, pkohvaei <address@hidden> wrote:
> Hello everybody,
>
> I have got some motion signals and I want to eliminate trivial moves which
> behave like noise for my whole digrams, I first tried a lowpass filter but
> it does not seem to work desirably since I'd rather need a means for
> eliminating data points smaller than a certein amplitude. I had a look into
> the other functions in the 'signal' package and fft looked promising. But
> again, do you have any experience in this , so with what functions would you
> go ahead?
>
> Cheers,
>
>
>
> --
> View this message in context: 
> http://octave.1599824.n4.nabble.com/Filtering-Signals-according-to-an-Amplitude-Range-tp4647610.html
> Sent from the Octave - General mailing list archive at Nabble.com.
> _______________________________________________
> Help-octave mailing list
> address@hidden
> https://mailman.cae.wisc.edu/listinfo/help-octave

If you filter base don amplitude you will also remove data points. If
in any case that is what you want to do. Select the points below the
threshold amplitude
below = signal <= Threshold
data   = signal(~below)

This doesn't sound like something you want to do anyways.

Check out the sgolayfilter function, those are shape preserving
filters that may work better than a simple lowpass.

You could also extract a "pattern" signal using matching filters, but
I do not know if that applies in your case.


Give us more detail, and we may be able to help you better.