How to get the peak value

[alex.]

May I know how to get the 2 peak values v1 and v2, and 2 time durations t1 and t2.

Since there is a lot small amplitude noise, difficult to use the Max function simplely.

May need to use some filter such as wavelet or FIR to denoise it first. May I know how to use them, any toolbox in labview. How to make the curve smooth first.

Thanks.

[jcarmody]

If you'd attach your data we (those of us that don't know the answer immediately) could have something to work with.

[alex.]

Please see the data attached.

[MattBradley · Accepted solution]

I can't look at your vi now (I'm at work), but if your signals are long enough you may want to consider a median filter.  You can set the number of points before and after to use.  I have found it to be useful for some (not all) smoothing issues.

[smercurio_fc]

The graph on your VI doesn't look anything like the graph in your image. You image is quite "jagged". There is also no discernible "second" peak. Which graph is more like what you actually have?

[alex.]

Although it looks smooth, it actually has over 200 peaks when using (x2-x1)*(x3-x2)<0 to calculate

[alex.]

Is the median filter inside some toolkit, thanks.

[smercurio_fc]

---

turbot wrote:

Although it looks smooth, it actually has over 200 peaks when using (x2-x1)*(x3-x2)<0 to calculate

---

I don't quite get what that calculation is supposed to prove, and I don't see 200 peaks. I see maybe 2, and that's dependent on how you define "peaks".

Median Filter - You need the Full Development version of LabVIEW.

[alex.]

With my eye, also detect 2 peaks.

But with the program and algorithm, there may be some problem.

Attached my testing program.

[F._Schubert]

I don't get resonable data out of the 23 file.

But an idea for this:

* Use a numerical derivative function followed by a trigger for the zero crossing. Thaht shoul give you some local maxima.

Is this related to your other post?

Felix