Abstract:Rhizome Chinese medicine refers to rhizomes or underground stem medicinal materials with a small amount of roots or fleshy scales. Due to the high similarity of plant morphology of rhizome medicinal materials,it has caused the problem of confusion and difficulty in distinguishing the authenticity of Chinese medicinal materials in the market,and further cause serious clinical medication risk.Traditional Chinese medicine identification methods are cumbersome to operate,long analysis time,high cost,and relatively low efficiency.In order to establish the identification method of six kinds of easily confused rhizome Chinese medicinal materials including Subprostrate Sophora,Ardisia crispa,Rhizoma Menispermi,Philippine Flemingia,Yunnan bean,and Yunnan cowpea,etc,which has rich chemical composition and high medicinal value. In this research we used near-infrared diffuse reflectance(NIR) spectroscopy to collect spectral information within 900-1 700 nm,and then combined with principal component analysis (PCA),systematic cluster analysis (SCA),k-nearest neighbor (k-NN),and linear discriminant analysis (LDA) to develop qualitative discriminant models. Results showed that the six kinds of medicinal materials showed obvious classification and aggregation characteristics in principal component analysis and systematic cluster analysis.,the classification accuracy rate for 46 unknown samples up to 93.48% and 95.65% for k-NN and LDA,respectively. From the results of the experiment,we come to the following conclusion that NIR coupled with pattern recognition techniques,which is rapid,accurate,and low-cost,has potential to discriminate Rhizome Chinese medicine.