A Novel Algorithm for Baseline Correction of Chemical Signals

As important as noise problem, baseline drift is another part for de-noising the signals acquired by contemporary measurement, especially in the field of chemical signals processing. A novel algorithm named Iterative Suppression Polynomial Fitting algorithm (ISPF) based on modified polynomial fitting method was proposed in this work, which could eliminate the baseline automatically compared to former mathematical methods. The principle of ISPF is intelligible. Raman spectra signal was selected as the investigated subject of experimental section. The drift baseline which is caused by fluorescence blocked up further analysis. After processing with ISPF algorithm, baseline drifts were subtracted from the original Raman spectra signals, rice grains from different places were clearly classified by Principal Component Analysis (PCA). The results proved the efficiency of ISPF algorithm, which could be extended to other field for signal de-noising.