Related Experiment Video
Updated: Dec 12, 2025
08:20
Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
Published on: October 27, 2023
2.3K
Grape Leaf Disease Identification Using Improved Deep Convolutional Neural Networks.
Bin Liu1,2,3, Zefeng Ding1, Liangliang Tian1
1College of Information Engineering, Northwest A&F University, Yangling, China.
Frontiers in Plant Science
|August 8, 2020
Summary
This study introduces a new deep learning model, DICNN, for identifying common grape leaf diseases like anthracnose and black rot. The model achieved 97.22% accuracy, improving grape disease diagnosis.
Area of Science:
- Agricultural Science
- Computer Science
- Plant Pathology
Background:
- Grape leaf diseases such as anthracnose, brown spot, mites, black rot, downy mildew, and leaf blight cause significant economic losses.
- Accurate and timely identification of these diseases is crucial for effective management and the healthy development of the grape industry.
Purpose of the Study:
- To propose a novel recognition approach for diagnosing grape leaf diseases using improved convolutional neural networks (CNNs).
- To develop a deep learning model capable of accurately identifying multiple common grape leaf diseases.
Main Methods:
- A large dataset of 107,366 grape leaf images was created using image enhancement techniques from field and public sources.
- An improved CNN architecture, termed DICNN, was developed, incorporating Inception structures for feature extraction and dense connectivity for feature reuse.
- The DICNN model was trained from scratch and evaluated on a hold-out test set.
Main Results:
- The proposed DICNN model achieved an overall accuracy of 97.22% in recognizing grape leaf diseases.
- DICNN demonstrated superior performance compared to GoogLeNet (2.97% increase) and ResNet-34 (2.55% increase).
- Experimental results confirm the model's efficiency in recognizing various grape leaf diseases.
Conclusions:
- The developed DICNN model offers an efficient and accurate method for diagnosing grape leaf diseases.
- This study provides a new approach for rapid plant disease diagnosis, laying a theoretical foundation for deep learning in agricultural information systems.
