Détection multimodale avancée et fusion de données pour la détection numérique précoce des symptômes de stress des plantes – MULTIFUSE

Climate change poses significant challenges to agriculture, intensifying extreme weather events and altering plant pathogen dynamics. Simultaneously, sustainable practices require reduced chemical pesticide and fertilizer use. Advanced sensing systems are needed to supplement traditional crop inspection, enabling precise individual plant treatment through efficient phenotyping and monitoring. Ensuring food security and sustainable agriculture requires early, accurate detection and differentiation of plant stress conditions. Current standards like multi- and hyperspectral imaging often fail to reliably link disease symptoms to specific stressors, limiting their utility for continuous crop monitoring. MULTIFUSE aims to integrate expertise from plant sciences, optical sensor technology, and data processing to exploit advanced multisensory systems for plant status evaluation. A major barrier in precision farming is the lack of integrated optical sensing data, preventing full utilization of potential synergies from digitization and complementary sensor data. MULTIFUSE seeks to revolutionize plant phenotyping through advanced multimodal optical sensing and data fusion. This data fusion from multiple sensors will create a digital system for detecting and differentiating plant stresses. Leveraging international expertise, MULTIFUSE aims to streamline the multimodal experimental workflow between the project partners and integrate advanced sensing technologies, including Raman spectroscopy, multichannel fluorescence imaging, optical coherence tomography, polarimetry, and imaging spectroscopy, for comparable phenotyping results. In the first phase, the project focuses on early detection and differentiation of abiotic stress factors like drought, water, and salt stress; representing omnipresent stressors in agriculture due to climate change. In the second phase and beyond the project duration, these technologies will also improve the detection of various plant diseases and pathogens, relevant to specific national interests in Europe and Japan. The project’s objectives are: 1. Establish a comparable screening platform for different sensing technologies across partner institutions. 2. Implement a joint (meta) data management system and a data handling platform for efficient cross-border sensor data exchange. 3. Develop advanced multimodal sensing approaches for detecting plant status and stress symptoms. 4. Establish digital data fusion models incorporating machine learning to identify and differentiate plant stress symptoms and potential pathogen infestations. MULTIFUSE’s findings will provide a valuable database, enabling the selection of optimal methods and promoting the translation of these technologies into practical agricultural environments.