Early Detection and Classification of Biotic and Abiotic Stress in Tomato Leaves Using AI-based Approaches-Review

The Tomato crops are highly susceptible to various biotic and abiotic stresses, significantly affecting yield and quality. Early detection and classification of these stresses are crucial for effective crop management and sustainable agriculture. This review explores the latest AI-based approaches for identifying and differentiating stress factors in tomato leaves, leveraging deep learning techniques such as Convolutional Neural Networks (CNNs), Vision Transformers (ViTs), and hybrid architectures. The study highlights the role of advanced image processing and spectral analysis in stress classification, emphasizing the use of the GKVK dataset for real-world applicability. This survey examines cutting-edge methods, including transfer learning, generative adversarial networks (GANs) for data augmentation, and multimodal fusion of image and sensor data. The paper also discusses emerging self-supervised learning techniques that enhance model performance with limited labeled data. Real-time and edge AI-based solutions for in-field deployment are explored, enabling precision agriculture applications. A comparative analysis of state-of-the-art models is provided, focusing on classification accuracy, robustness, and scalability. The review concludes with key challenges, including dataset diversity, model interpretability, and ethical considerations in AI-driven agriculture. This work serves as a comprehensive resource for researchers and practitioners, paving the way for AI-powered smart farming solutions.