Extraction of Chromatographic Characteristics Based on Wavelet Ridge and Morphology

In order to improve the quality of peak characteristics extraction in chromatographic data, an automatic peak fitting method based on wavelet ridges and morphology (WMPF) was proposed. In handling overlapping peaks and weak peaks, WMPF gave solution to deal with the shortcomings of existing peak detection methods. The ridge line of the signal was obtained by wavelet transform of the original chromatogram signal, and the signal was divided into peak areas and non-peak areas. Then, we processed morphological operations with an adaptive window, and combined Gaussian fitting with wavelet ridge information. Finally, the peak characteristics of the signal were extracted. Through the analysis of high-density lipoprotein electrophoretogram data, it was confirmed that the proposed method was effective for baseline separation, noise elimination, peak detection and overlapping peak recognition.