Recognition effect of models based on different microscope objectives

Abstract

Lonicerae japonicae flos, a common clinical Chinese medicine, is widely used in proprietary traditional Chinese medicine for the treatment of various conditions, such as fever, cough, and influenza. The microscopic features of honeysuckle pollen grains significantly correlate with their medicinal effects. In this study, deep learning using artificial intelligence was cross-combined with microscopic images of Chinese herbal medicines, and we proposed microscopic identification through an intelligent recognition method of honeysuckle pollen grains using microscopic images based on YOLO v5. The expandability of the microscopic feature recognition of different magnification models was verified based on different microscopic objectives. The honeysuckle pollen grains model based on YOLO v5 can quickly and accurately identify the microscopic images of pollen grains, which can provide a reference for the quality improvement and quality standardization of traditional Chinese herbs and has good application prospects.