Related Experiment Video
Updated: Jul 11, 2025
08:27
Author Spotlight: Efficient Image Recognition Using Directional Gradient Histogram Technique and Support Vector Machines
Published on: January 5, 2024
1.1K
Dried shiitake mushroom grade recognition using D-VGG network and machine vision.
Li Wang1, Penghao Dong1, Qiao Wang1,2
1School of Electrical Engineering, Henan University of Technology, Zhengzhou, China.
Frontiers in Nutrition
|November 3, 2023
Summary
This study introduces an advanced machine vision method for grading dried shiitake mushrooms, achieving 96.21% accuracy. The developed D-VGG model efficiently recognizes mushroom grades, improving quality control in agricultural products.
Area of Science:
- Agricultural Science
- Computer Vision
- Machine Learning
Background:
- Dried shiitake mushroom grading is crucial for marketability due to quality variations.
- Automatic grade recognition faces challenges due to irregular shapes and subtle morphological differences.
Purpose of the Study:
- To develop a comprehensive machine vision method for accurate and efficient dried shiitake mushroom grade recognition.
- To address the limitations of existing methods in handling irregular shapes and mixed grades.
Main Methods:
- Image acquisition, preprocessing, and dataset creation.
- Osprey Optimization Algorithm (OOA) for efficient Otsu's thresholding and contour extraction.
- A novel D-VGG network integrating VGG16, residual modules, batch normalization, and channel attention for enhanced feature learning.
Main Results:
- The D-VGG model achieved a high grading accuracy of 96.21%.
- Processing a single image took only 46.77 ms, demonstrating high recognition efficiency.
- The proposed method effectively handles challenges posed by irregular shapes and subtle grade differences.
Conclusions:
- The D-VGG model offers a robust and efficient solution for dried shiitake mushroom quality grading.
- This approach provides a valuable reference for real-time grade recognition of other agricultural products.
