A Ginseng Appearance Quality Grading Method Based on an Improved ConvNeXt Model

A Ginseng Appearance Quality Grading Method Based on an Improved ConvNeXt Model

In order to solve the problem of the small degree of variability between the features of ginseng grading classes and the resulting need for heavy reliance on professionals, this study established a ginseng dataset containing 5116 images with three classes in different contexts and proposed a ginseng...

Full description

Bibliographic Details
Main Authors: Dongming Li, Mengting Zhai, Xinru Piao, Wei Li, Lijuan Zhang
Format: Article
Language:English
Published: MDPI AG 2023-06-01
Series:Agronomy
Subjects:
deep learning
ginseng grading
ConvNeXt
Channel Shuffle
structural re-parameterization
activation function
Online Access:https://www.mdpi.com/2073-4395/13/7/1770

Internet

https://www.mdpi.com/2073-4395/13/7/1770

Similar Items