Reputation: 2091
I have a raw ecg signal, that contains complex values (real and imaginary) in time. Now I have to clear that signal out, remove noises, and flatten the signal.
The algorithm to do this that i know of is fast fourier transformation (FFT), but it doesnt flatten the signal, instead it generates correct fourier transformation, but the signal is not flat, it has high values on both sides. How can i do that?
I am doing this in java language, but I dont ask for the code, just for the hint with the idea, or an algorithm.
Thanks!
Upvotes: 0
Views: 95
Reputation: 80197
FFT doesn't flatten signal, it translates signal from time domain to frequency domain. If you signal is pure real, FT is symmetric - so you can see similar high peaks at both ends - this is very low frequency part of your signal.
To filter a signal, you can execute FT, apply some function to the result of transform - for example, lower high and very low frequency regions, and execute backward FT to return in the time domain.
Upvotes: 1