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

2024-05-30

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.

Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

This paper aims to address an issue in plant disease detection, specifically focusing on the accurate and efficient identification of tomato leaf diseases. Machine learning has played a significant role in agricultural science, particularly in early plant disease management and detection. The paper proposes an innovative method based on Convolutional Neural Networks (CNN), combining Tensor Subspace Learning with High-Order Whitening Singular Value Decomposition (HOWSVD) and Multilinear Discriminant Analysis (MDA). This method improves the classification of tomato leaf diseases by enhancing the discriminative ability of the system. The paper first introduces the use of pre-trained CNN models to extract features and enhance the discriminative power of these features through HOWSVD. Then, tensor subspace learning is implemented through a two-stage approach, which involves HOWSVD followed by MDA. Experimental results demonstrate that this method outperforms existing methods on two different datasets (PlantVillage and a dataset from Taiwan), achieving accuracies of up to 98.36% and 89.39%, respectively. The main contributions of the paper include: 1. Conducting a literature review to identify research gaps. 2. Proposing a novel tensor subspace learning method called HOWSVD for the classification of tomato leaf diseases, which improves the system's discriminative ability by learning discriminative features from high-dimensional CNN embeddings. 3. Implementing a two-step subspace projection process using HOWSVD-MDA, which combines unsupervised HOWSVD with supervised MDA to enhance the system's discriminative ability. 4. Conducting a comprehensive comparative analysis of multiple pre-trained CNN models to determine the most suitable model for a specific classification task. 5. Conducting extensive experiments on two datasets to validate the effectiveness of the proposed HOWSVD-MDA method and its robustness for real-world applications.

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

Crop Disease Identification by Fusing Multiscale Convolution and Vision Transformer.

Deep Learning for Tomato Diseases: Classification and Symptoms Visualization

Precision Agriculture Through Deep Learning: Tomato Plant Multiple Diseases Recognition With CNN and Improved YOLOv7

Detection and classification of plant diseases in leaves through machine learning

Synergistic use of handcrafted and deep learning features for tomato leaf disease classification

Early Detection and Classification of Tomato Leaf Disease Using High-Performance Deep Neural Network

Impact of Pretrained Deep Neural Networks for Tomato Leaf Disease Prediction

On the Image-Based Detection of Tomato and Corn leaves Diseases : An in-depth comparative experiments

A detection of tomato plant diseases using deep learning MNDLNN classifier

Design of Efficient Methods for the Detection of Tomato Leaf Disease Utilizing Proposed Ensemble CNN Model

Transform and Deep Learning Algorithms for the Early Detection and Recognition of Tomato Leaf Disease

Plant disease management: a fine-tuned enhanced CNN approach with mobile app integration for early detection and classification

Tomato Plant Leaf Disease Detection Using Convolutional Neural Network

Tomato Leaf Disease Classification via Compact Convolutional Neural Networks with Transfer Learning and Feature Selection

A robust deep learning approach for tomato plant leaf disease localization and classification

Detection and Classification of Tomato Crop Disease Using Convolutional Neural Network

CSXAI: a lightweight 2D CNN-SVM model for detection and classification of various crop diseases with explainable AI visualization

Agricultural innovation through deep learning: a hybrid CNN-Transformer architecture for crop disease classification

Tomato plant leaf diseases detection and classification using an advanced self-developed CNN model

Multimodal Hybrid Deep Learning Approach to Detect Tomato Leaf Disease Using Attention Based Dilated Convolution Feature Extractor with Logistic Regression Classification