Filtering and smoothing data from Infrared Thermography

Hello everyone.

I am dealing with third order tensor which is a set of raw results from infrared thermography. The structure of the data is the following.

1. dimension one = pixel
2. dimension two = pixel
3. dimenstion three = time (time steps)

I need to apply some de-trending (smoothing the data), I have checked that the level of noise in the data is too high to apply any smoothing directly to the data. Therfore, I am thinking to apply some filter (low pass, broad band or high pass). Since I am very novel to applying filters to thermographic data, I was wondering if someone can enlight me more about the filtering process. In particular, I am interested in knowing which filter has usable (also not-usable) and possibly know more about which are the best parameters.

For you information: These data were acquired at a sampling frequency of 10Hz (10 Frames per second), and the target sample was subject to vibration applied by a piezoelectric disk glued on the surface of the sample.

For any further elucidatio please do not hesitate to ask.

Have a good day.

Luca