peak tracking across multiple signals

[eword]

Peak tracking across multiple signals

I have a series of PSD graphs which represent the acoustic information of an object at several different angles. What I am trying to
do is track three important peaks out of these signals, and calculate their height and width. I understand how to use the peak detector .vi
and I can locate all the peaks in each signal and calculate these parameters, however, I am unsure how I can isolate the peaks that
I am interested in. To understand what I am please see the pictures attached.

I did think of putting cursors on my graph and searching for the max peak between the cursors in the regions where each of these peaks lie, however, given the peaks move a lot along
the x-axis a lot with each angle I am not sure if this is the optimum solution. Are there any other methods I could be using?

[eword]

Here is my peak finding .vi as a reference.

[Andy_C.]

See this post

[eword]

I gathered I could do something like this, however, what happens if the peak that I am tracking disappears, or suddenly divides into several different peaks. Then it will jump to the next peak, which I am not interested in.

[eword]

Here is an illustration of what I think the .vi you sent me will do, which is not what I want. 1,2,3 represent the peaks that it will find, but the red zones indicate the peaks I am interested in.

[jareds]

Are you trying to only identify the highest peak in the cluster?  Or are you trying to identify these three clusters themselves?  If you're just looking to identify the clusters, you could filter the data to smooth out some of the peaks.  Then you could check for certain thresholds to identify a cluster from the surrounding data.  In this way you could identify the clusters.

If you're looking for the peaks within each cluster, you could then use the clusters that you have found to find the highest peak in each data range.