I'm very new to signal processing. I have some data measured over time at 800 Hz. I'm interested in the high frequency signals. You can easily identify them in the image below (in this case, 3 signals):
You can easily see that there is a visible trend in the noise. This trend is mostly different for every such data that I have. I know that this is due to the environmental effects (temperature) but unfortunately, I don't have a temperature sensor set up.
I'm looking for a scientific way to extract these signals from the noisy data. Currently, I visually find the beginning and the end of each signal, consider the beginning of the signal to be zero and then offset the entire signal by that value.
Is there a way through which I can process the entire data and remove the noise from it entirely in one go? Are there any built in functions in MATLAB for the same?