Re: determining CFAR scaling factor
>What are you scaling? You should get a pretty uniform behavior if you
>scale the mean, or median, or some other statistic of the signal. So if
you
>set the threshold to c*m, where m is the moving average of some length,
you
>should get the same peaks on Mac or Windows (assuming the signals differ
>only by a constant!)
Above, c is the scale factor, same in either case (Mac or Windows). Just
to clarify...
Emre
|