Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Enhancing Plant Disease Detection: A Novel CNN-Based Approach with Tensor Subspace Learning and HOWSVD-MD

Abdelmalik Ouamane, Ammar Chouchane, Yassine Himeur, Abderrazak Debilou, Abbes Amira, Shadi Atalla, Wathiq Mansoor, Hussain Al Ahmad

TL;DR

This work tackles tomato leaf disease detection by fusing deep features from multiple pre-trained CNNs into a tensor framework and applying a two-stage HOWSVD-MDA subspace learning chain. The method whitened higher-order SVD (HOWSVD) followed by Multilinear Discriminant Analysis (MDA) yields compact, discriminative tensor embeddings, enabling robust classification via cosine similarity. Across PlantVillage and Taiwan tomato datasets, the approach with four CNN models achieves state-of-the-art accuracy, up to 98.36% and 98.39% respectively, demonstrating the advantage of high-order tensor representations and knowledge-driven feature fusion. The results suggest strong practical potential for scalable, multi-view plant disease diagnostics in agricultural settings and IoT-enabled systems, with opportunities for extension to federated learning contexts.

Abstract

Machine learning has revolutionized the field of agricultural science, particularly in the early detection and management of plant diseases, which are crucial for maintaining crop health and productivity. Leveraging advanced algorithms and imaging technologies, researchers are now able to identify and classify plant diseases with unprecedented accuracy and speed. Effective management of tomato diseases is crucial for enhancing agricultural productivity. The development and application of tomato disease classification methods are central to this objective. This paper introduces a cutting-edge technique for the detection and classification of tomato leaf diseases, utilizing insights from the latest pre-trained Convolutional Neural Network (CNN) models. We propose a sophisticated approach within the domain of tensor subspace learning, known as Higher-Order Whitened Singular Value Decomposition (HOWSVD), designed to boost the discriminatory power of the system. Our approach to Tensor Subspace Learning is methodically executed in two phases, beginning with HOWSVD and culminating in Multilinear Discriminant Analysis (MDA). The efficacy of this innovative method was rigorously tested through comprehensive experiments on two distinct datasets, namely PlantVillage and the Taiwan dataset. The findings reveal that HOWSVD-MDA outperforms existing methods, underscoring its capability to markedly enhance the precision and dependability of diagnosing tomato leaf diseases. For instance, up to 98.36\% and 89.39\% accuracy scores have been achieved under PlantVillage and the Taiwan datasets, respectively.

Enhancing Plant Disease Detection: A Novel CNN-Based Approach with Tensor Subspace Learning and HOWSVD-MD

TL;DR

This work tackles tomato leaf disease detection by fusing deep features from multiple pre-trained CNNs into a tensor framework and applying a two-stage HOWSVD-MDA subspace learning chain. The method whitened higher-order SVD (HOWSVD) followed by Multilinear Discriminant Analysis (MDA) yields compact, discriminative tensor embeddings, enabling robust classification via cosine similarity. Across PlantVillage and Taiwan tomato datasets, the approach with four CNN models achieves state-of-the-art accuracy, up to 98.36% and 98.39% respectively, demonstrating the advantage of high-order tensor representations and knowledge-driven feature fusion. The results suggest strong practical potential for scalable, multi-view plant disease diagnostics in agricultural settings and IoT-enabled systems, with opportunities for extension to federated learning contexts.

Abstract

Machine learning has revolutionized the field of agricultural science, particularly in the early detection and management of plant diseases, which are crucial for maintaining crop health and productivity. Leveraging advanced algorithms and imaging technologies, researchers are now able to identify and classify plant diseases with unprecedented accuracy and speed. Effective management of tomato diseases is crucial for enhancing agricultural productivity. The development and application of tomato disease classification methods are central to this objective. This paper introduces a cutting-edge technique for the detection and classification of tomato leaf diseases, utilizing insights from the latest pre-trained Convolutional Neural Network (CNN) models. We propose a sophisticated approach within the domain of tensor subspace learning, known as Higher-Order Whitened Singular Value Decomposition (HOWSVD), designed to boost the discriminatory power of the system. Our approach to Tensor Subspace Learning is methodically executed in two phases, beginning with HOWSVD and culminating in Multilinear Discriminant Analysis (MDA). The efficacy of this innovative method was rigorously tested through comprehensive experiments on two distinct datasets, namely PlantVillage and the Taiwan dataset. The findings reveal that HOWSVD-MDA outperforms existing methods, underscoring its capability to markedly enhance the precision and dependability of diagnosing tomato leaf diseases. For instance, up to 98.36\% and 89.39\% accuracy scores have been achieved under PlantVillage and the Taiwan datasets, respectively.
Paper Structure (23 sections, 11 equations, 9 figures, 9 tables)

This paper contains 23 sections, 11 equations, 9 figures, 9 tables.

Table of Contents

  1. Introduction
  2. Related works
  3. HOWSVD-MDA Subspace Learning
  4. Materials and Formulations
  5. Tensor definition
  6. k-Mode Unfolding (Matricization)
  7. k-Mode Product of a tensor
  8. Higher-Order Whitened Singular Value Decomposition (HOWSVD)
  9. HOWSVD-MDA Algorithm
  10. Knowledge pre-trained CNN based tensor subspace learning
  11. Knowledge Pre-trained CNN
  12. Tensor subspace learning
  13. Matching and Classification Decision
  14. Experiments and Results Analysis
  15. Datasets Description
  16. ...and 8 more sections

Figures (9)

  • Figure 1: Samples of tomato leaves images from PlantVillage dataset
  • Figure 2: Plant disease detection and classification based-methods
  • Figure 3: Visualization of the 3-Mode Unfolding of a 3$^{th}$-order tensor
  • Figure 4: Block diagram of the proposed system
  • Figure 5: Knowledge-based multi Pre-trained CNN features extraction and tensor design for tomato leaves images
  • ...and 4 more figures