Hello dear, you can use a signal in your spectra whose intensity remain stable or its variation be smallest (this signal appear at low frequency, usually at wave number lower than 1500 cm-1 , and then you do the relation of the intensity between the signals of your interest and this signal.
If you measure films, your task is definitely non-trivial as the signals are enhanced or weakened by interference. Furthermore, band strengths does not depend linearly on the film thickness. Instead, there is a kind of resulting thickness that depends on the optical constants. If this is relevant for your samples you can find details in
My paper titled "determination of OH-nmber by using of FTIR spectroscopy " that is published in "Polymer Testing" Journal ; Elsevier could be useful for your studies.