From: Matt J on 13 Jul 2010 12:35 "adi cohen" <kerenpur(a)yahoo.com> wrote in message <i1htdm$a1a$1(a)fred.mathworks.com>... > I want to assum that the signal is WSS and that the noise is added, WSS as well. But the wiener filter ruins the shape of the my signal...is there a filter that does a better job at maintaining the shape? =============== If you know something more about the signal than just the fact that it is WSS, you might be able to do a more sophisticated estimation. For example, if your signal has a known parametrized shape, you could pose this as a curve fitting problem, and use the Curve Fit Toolbox, for example.
From: Greg Heath on 14 Jul 2010 08:09 On Jul 13, 10:32 am, "adi cohen" <keren...(a)yahoo.com> wrote: > "Matt J " <mattjacREM...(a)THISieee.spam> wrote in message <i1hqdm$p7...(a)fred.mathworks.com>... > > > "adi cohen" <keren...(a)yahoo.com> wrote in message <i1hodp$in...(a)fred.mathworks.com>... > > > > In order to restore it but maintain the shape of the signal as much as possible, what is the best approach? > > ============== > > > That really depends on what you know in advance about the signal and the process that corrupts it. Since you're considering a wiener filter, it presumably means the signal is WSS and the corruption of the signal is additive noise, also WSS. > > > > I thought about using a wiener filter, but cannot find a function in matlab > > > that filters one-dimentional signals using the wiener algorithm > > ================= > > > If you have the Image Processing Toolbox, you could probably use deconvwnr(). Even if your signal is 1-dimensional, you could just pretend that it's a 1-dimensional image. > > thank you, > I want to assum that the signal is WSS and that the noise is added, WSS as well. But the wiener filter ruins the shape of the my signal...is there a filter that does a better job at maintaining the shape? Try posting in comp.dsp. Hope this helps. Greg
From: adi cohen on 28 Jul 2010 02:58 thank you for your help for some reason, there aren't any functions in matlab dealing with restoration of one-dimentional signals. Treating the signal as an image in order to use written function in matlab only complicates my code.
|
Pages: 1 Prev: using fscanf to read a datafile Next: Error starting matlab on windows 7 |