Related Experiment Video
Updated: Aug 6, 2026
11:48
A Contrast of Three Inoculation Techniques used to Determine the Race of Unknown Fusarium oxysporum f.sp. niveum Isolates
Published on: October 28, 2021
Advancing sustainable agriculture through multi-parameter fuzzy soft set-based plant disease classification
D Rajalakshmi1, K Kannan2, A Menaga3
1Department of Mathematics, Srinivasa Ramanujan Centre, SASTRA Deemed to be University, Kumbakonam, Tamil Nadu, 612001, India. drajilak@gmail.com.
Scientific Reports
|July 18, 2026
Summary
This study introduces an interpretable fuzzy soft set framework for classifying tomato plant diseases from leaf images. The model shows improved performance in cross-dataset validation, outperforming other classifiers in uncertain conditions.
Area of Science:
- Agricultural Science
- Computer Science
- Data Science
Background:
- Plant diseases pose significant threats to agricultural productivity and global food security.
- Accurate identification of plant diseases from leaf images is challenging due to overlapping visual symptoms and uncertainty.
- Existing deep learning models often require large datasets and lack interpretability in real-world agricultural settings.
Purpose of the Study:
- To develop an interpretable and reliable plant disease classification framework using an improved fuzzy soft set approach.
- To handle uncertainty in plant disease patterns through feature-driven fuzzy similarity analysis.
- To classify tomato plant diseases using leaf images from the PlantVillage dataset.
Main Methods:
- Image preprocessing and extraction of color (RGB, HSV) and texture (GLCM) features.
- Variance-based feature weighting, K-means clustering for prototype generation, and fuzzy similarity computation (Mahalanobis distance, Gaussian membership functions).
- Comparison with Support Vector Machine (SVM), Random Forest (RF), Linear Discriminant Analysis (LDA), and Naive Bayes (NB) classifiers using Python Scikit-learn.
Main Results:
- The proposed Improved Fuzzy Soft model achieved 88.57% accuracy on the primary dataset, lower than SVM (97.94%), RF (97.78%), and LDA (94.92%).
- In cross-dataset validation, the Improved Fuzzy Soft model achieved 67.35% accuracy, outperforming LDA (51.02%), RF (51.02%), and SVM (55.10%).
- Statistical analysis (Wilcoxon Signed-Rank Test) indicated a significant performance difference between the proposed framework and Random Forest in cross-dataset validation.
Conclusions:
- The Improved Fuzzy Soft framework offers competitive classification performance with enhanced interpretability and robustness, particularly in handling uncertain feature distributions.
- The model demonstrates superior generalization capabilities in cross-dataset validation, highlighting its potential for practical agricultural applications.
- The study underscores the effectiveness of fuzzy set theory in addressing the inherent uncertainties in plant disease identification from visual data.
