This work presents a new spectral data compression-rebuilding technique to
translate the full IR spectral data into compact codes based on the analysi
s and comprehension encoding approach. This method has been successfully ap
plied to a sample set of 505 IR spectra randomly picked from 100 000 spectr
a. The results show that the compression ratio reaches 12:7:1 under a very
weak curve distortion. The choice of the number and shape of the basis func
tions is flexible. The IR spectra can be compressed in a fixed data size in
fulfilling the distortion criteria. The data after compression have no sig
nificance in the sense of IR spectra. To recover the original spectra, a sp
ecific algorithm must be applied. So the method can be used as a cryptic to
ol. Furthermore, the method can be applied to the compression of other comp
lex curve by utilizing some of proper basis functions.