Related Experiment Video
Updated: Jul 12, 2025
06:41
Imaging and Analysis for Quantifying Maize (Zea mays) Abiotic Stress Phenotypes
Published on: March 28, 2025
868
Advanced deep learning techniques for early disease prediction in cauliflower plants.
G Prabu Kanna1, S J K Jagadeesh Kumar2, Yogesh Kumar3
1School of Computer Science and Engineering, VIT Bhopal University, Bhopal-Indore Highway, Kothrikalan, Sehore Madhya Pradesh - 466114, India.
Scientific Reports
|October 27, 2023
Summary
This study uses deep transfer learning to accurately identify cauliflower diseases, improving agricultural sustainability and food security. Advanced models like EfficientNetB1 show high accuracy in detecting issues like bacterial spot rot and downy mildew.
Area of Science:
- Agricultural Science
- Computer Vision
- Machine Learning
Background:
- Crop diseases threaten food security and farmer livelihoods in developing nations.
- Limited studies exist for cauliflower disease identification due to data and surveillance challenges.
- Accurate disease detection is crucial for effective agricultural management.
Purpose of the Study:
- To enhance cauliflower disease identification and detection in rural agriculture using deep transfer learning.
- To address challenges of insufficient disease surveillance and limited high-quality datasets.
- To explore the significance of automated cauliflower disease classification.
Main Methods:
- Trained and evaluated ten deep transfer learning models (EfficientNetB0-B4, Xception, MobileNetV2, DenseNet201, InceptionResNetV2, ResNet152V2) on four cauliflower disease classes from the VegNet dataset.
- Assessed model performance using metrics including accuracy, loss, precision, recall, and F1-score.
- Focused on Bacterial spot rot, Black rot, Downy Mildew, and healthy cauliflower classifications.
Main Results:
- EfficientNetB1 achieved the highest validation accuracy (99.90%), lowest loss (0.16), and lowest root mean square error (0.40).
- All evaluated deep transfer learning models demonstrated strong capabilities in classifying cauliflower diseases.
- The study successfully classified four distinct cauliflower disease categories.
Conclusions:
- Advanced Convolutional Neural Network (CNN) models are critical for automating cauliflower disease detection and classification.
- These automated systems can lead to robust applications for disease management, benefiting farmers and consumers.
- Deep transfer learning offers a powerful solution to improve agricultural surveillance and protect crop yields.
