Study on the Extraction Method of Glycyrrhiza uralensis Fisch. Distribution Area Based on Gaofen-1 Remote Sensing Imagery: A Case Study of Dengkou County

Xinxin Wei ^1,2 Ze Yuan Zhao ³ Taiyang Chen ¹ Xiaobo Zhang ² Shuying Sun ^1* Minhui Li ³ Tingting Shi ²

• ¹ School of Life Science, Inner Mongolia University, Hohhot, China
• ² State Key Laboratory of Dao-di Herbs, National Resource Center for Chinese Materia Medica, China Academy of Chinese Medical Sciences, Beijing, China
• ³ Inner Mongolia Traditional Chinese & Mongolian Medical Research Institute, Hohhot, Inner Mongolia Autonomous Region, China

Glycyrrhiza uralensis Fisch., a perennial medicinal plant with a well-developed root system, plays a critical role in preventing land desertification when cultivated extensively. This study focuses on Dengkou County, a semi-arid region, as the research area. First, by analyzing the spectral characteristics of the Gaofen-1 Satellite images of the main feature types under different bands and months, the significance of the difference between the blue and the near-infrared band for identifying G. uralensis was clarified. Then, the importance of bands was assessed using the Random Forest (RF) algorithm, revealing that incorporating data from January to December significantly enhances the identification of G. uralensis. Secondly, after constructing the G. uralensis vegetation index (GUVI) based on the above results, the recognition accuracy of G. uralensis was compared between the RF classification model constructed based on the January-December GUVI and common vegetation indices feature set and the support vector machine (SVM) classification model constructed on the GUVI feature set. The results demonstrated that the RF classification model using the GUVI feature set achieved superior performance, with a producer's accuracy of 97.26%, an overall accuracy of 93.00%, a Kappa coefficient of 91.38%, and a user's accuracy of 97.32%. This indicates that the RF classification model based on GUVI is more efficient and robust than models using other feature sets or classification algorithms. Finally, the differences between the spectral features of G. uralensis and other feature types under the 2022 GUVI feature set were analyzed, and the historical distribution area of G. uralensis was successfully identified and extracted. It was found that the distribution area of G. uralensis was initially distributed in the northeast of Dengkou County from 2014 to 2022, and gradually distributed to the middle, and now it is also distributed in the southwest. The distribution area also gradually changed from the initial small and scattered to a large area of concentrated distribution. This study provides a reference for utilizing remote sensing technology to accurately identify current and historical G. uralensis distribution areas in regions with similar conditions to the study area.