MULTIMODAL ARTIFICIAL INTELLIGENCE FOR EARLY PLANT STRESS PREDICTION: A REVIEW
DOI:
https://doi.org/10.29121/SARPS.v2.i3.2026.48Keywords:
Artificial Intelligence, Plant Stress, Computer Vision, Precision AgricultureAbstract
Plant stress is one of the primary limitations to crop growth, yield, and quality and can result from various abiotic and biotic stresses, including drought, extreme temperatures, salinity, nutrient imbalance, pathogens, and insect pests. Early detection of plant stress is crucial for enabling timely intervention and preventing potential crop losses. In recent years, the use of artificial intelligence (AI) for automated plant health assessment has gained increasing attention through image-based deep learning and sensor-based machine learning techniques. However, single-modality approaches may have limited capability to capture the complex relationships between plant visual characteristics and environmental conditions. Multimodal AI can address this limitation by integrating complementary information from visual modalities, such as RGB, thermal, multispectral, and hyperspectral imaging, with environmental indicators, including temperature, humidity, soil moisture, light intensity, and weather conditions. In AI-based plant stress assessment, with a particular focus on multimodal data fusion and early stress prediction. AI architectures, data fusion methods, datasets, sensing modalities, and explainable AI (XAI) techniques are examined in detail. Furthermore, key challenges associated with multimodal dataset collection, sensor synchronization, environmental variability, model generalization, computational requirements, and real-world deployment are discussed. Finally, promising research directions, including temporal multimodal learning, edge AI, explainable multimodal models, low-cost sensing technologies, and foundation models for plant science, are highlighted.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 G.S. Deepti Sri, V.Lavanya Devi, S. Saravanan (Author)

This work is licensed under a Creative Commons Attribution 4.0 International License.
With the licence CC-BY, authors retain the copyright, allowing anyone to download, reuse, re-print, modify, distribute, and/or copy their contribution. The work must be properly attributed to its author.
It is not necessary to ask for further permission from the author or journal board.
This journal provides immediate open access to its content on the principle that making research freely available to the public supports a greater global exchange of knowledge.



















