Identification System Of Glycyrrhiza Seeds Based On Near Infrared Hyperspectral

Glycyrrhiza is a kind of genuine medicinal material with large demand and wide range of uses in China.Glycyrrhiza seeds,as an important means of Glycyrrhiza production,have attracted the attention of researchers of traditional Chinese medicine.In their researches,it is necessary to achieve the identification and sorting of Glycyrrhiza seeds.However,at present,the identification of Glycyrrhiza seeds' origin and other information by researchers still remains in morphological observation,chemical analysis and biological analysis.These three methods have some disadvantages,such as low accuracy,long identification period and damage of Glycyrrhiza seeds.In order to solve the shortcomings of the above identification methods and speed up the researches of researchers,this paper used the near infrared hyperspectral technology,which has developed rapidly in recent years and was widely used in the qualitative analysis of crops,developed a desktop application.The application can quickly and nondestructively identify the origin and variety information of Glycyrrhiza seeds.In this paper,the simple threshold segmentation and OTSU were used to extract the region of interest(ROI)of Glycyrrhiza seeds in hyperspectral images,so as to obtained the spectral information of Glycyrrhiza seeds.Different data processing and modeling methods were compared,K-means based down sampling method was used to reduce the data imbalance of the sample set;Sample Set Partitioning Based on Joint x-y Distance(SPXY)method was used to divide the sample set;Savitzky-Golay(SG)smoothing + Multiplicative Scatter Correction(MSC)+ First Derivative was used for spectrum preprocessing;Independent Component Analysis(ICA)was used for features extraction;Support Vector Machine(SVM)classifier was used to establish models for identifying the origin and variety of Glycyrrhiza seeds.According to the research results,the identification system was designed and developed.Chinese medicine researchers can use this system to realize the modeling of known Glycyrrhiza seeds and the identification of unknown Glycyrrhiza seeds.After testing,the effect was ideal,which met practical needs of researchers to identify and sort Glycyrrhiza seeds,and provided assistance for the research work of Glycyrrhiza seeds.