It has the application potentials of standoff detection of biological aerosol by passive infrared remote sensing based on Fouriertransform infrared (FTIR) spectroscopic technique. There are often noise signals and baseline drift on the target spectral signature in infrared remote sensing measurement. The spectrum of biological aerosol is relatively broad and the traditional methods of baseline correction are inapplicable. Due to infrared spectra of biological aerosol and different baseline drift are nonGaussian signals, an algorithm for preprocessing infrared spectra of biological aerosol is devised based on independent component analysis (ICA), where nonGaussian is used as independent measure. The results of experiments show that this algorithm can separate unknown interference and baseline drift as independent component, and with no effects on the further qualitative and quantitative analysis.