Fingerprint data extraction from Chinese herbal medicines with terahertz spectrum based on second-order harmonic oscillator model

Abstract

Terahertz time-domain spectroscopy provides a new method of identifying and studying Chinese herbal medicines. We utilize a second-order terahertz filter to describe the samples of Chinese herbal medicines, and the behaviors of terahertz filter are characterized by a second-order differential equations with three parameters {α, β, γ} or the corresponding frequency domain system function H(f). The results of data analysis show that five kinds of Chinese herbal medicines which are used in the experiments have unique filter parameters. Thus, if these filter parameters are used as the fingerprint data of Chinese herbal medicines, we can effectively determine the types and origins of Chinese herbal medicines.