DOI: 10.1002/cem.70182 ISSN: 0886-9383
Geographical Origin Tracing of
Dendrobium officinale
in Different Spatial Regions Based on a Multispectral Data Strategy Combined With Deep Learning
Guangyao Li, Zhili Duan, Yuanzhong Wang ABSTRACT
The environment in which plants are grown significantly affects their quality, particularly for species with both medicinal and edible value, such as
Dendrobium officinale
(
D. officianle
). This study proposes a method of tracking the geographical origin of
D. officinale
in different regions using a combination of 3D multispectral data and deep learning. Using 3D MIR and NIR spectral data, the study compares and analyzes the spatial representativeness of geographic information between single‐spectral and data fusion techniques. By integrating deep learning models, we have achieved 100% accuracy in tracking the geographical origin of
D. officinale
in different spatial regions. This research provides new technical means for tracking the origin of plants and ensuring the traceability of agricultural products, particularly Chinese herbal medicines with unique regional characteristics, such as
D. officinale
. It offers new ideas for origin tracing and quality assurance.