关键词: 光智农业, 数据感知技术, 光学传感器, 智能决策系统, 多源数据融合

Abstract:

[Significance] Opto-intelligent agriculture represents an emerging paradigm that deeply integrates optical sensing and intelligent decision-making within agricultural systems, aiming to transform production from experience-based management to data-driven precision cultivation. The core of this paradigm lies in exploiting the dual role of light: As an information carrier, it enables non-destructive sensing of crop physiological states through hyperspectral imaging, fluorescence, and other optical sensors; As a regulatory factor, it allows feedback-based manipulation of the light environment to precisely regulate crop growth. This establishes a closed-loop framework of "perception-decision-execution", which substantially enhances water and fertilizer use efficiency, enables early warning of pests and diseases, and supports quality-oriented production. Nevertheless, the transition of this technology from laboratory research to large-scale field application remains challenged by unstable signals under complex environments, weak model generalization, high equipment costs, a shortage of interdisciplinary talent, and insufficient policy support and promotion mechanisms. This paper systematically reviews the technological architecture, practical achievements, and intrinsic limitations of opto-intelligent agriculture, with the objective of providing theoretical guidance and practical directions for future development. [Progress] Opto-intelligent agriculture is evolving from isolated technological advances toward full-chain integration, characterized by significant progress in optical sensing, intelligent decision-making, and precision execution. At the optical sensing level, technological approaches have expanded from traditional spectral imaging to multi-scale, synergistic sensing networks. Hyperspectral imaging captures subtle spectral variations during the early stages of crop stress, chlorophyll fluorescence imaging enables ultra-early diagnosis of both biotic and abiotic stresses, LiDAR provides accurate three-dimensional phenotypic data, and emerging quantum-dot sensors have enhanced detection sensitivity down to the molecular scale. In terms of intelligent decision-making, recent advances focus on the deep integration of mechanistic and data-driven models, which compensates for the limited adaptability of purely mechanistic models while improving the interpretability of purely data-based ones. Through multi-source data fusion, the system jointly analyzes optical, environmental, and soil parameters to generate globally optimal strategies that balance yield, quality, and resource efficiency. At the execution stage, systems have developed into real-time feedback control loops. Dynamic light-spectrum LED systems and intelligent variable-spray drones transform decision outputs into precise actions, while continuous monitoring enables adaptive self-optimization. This mature technological chain has delivered measurable outcomes across the agricultural value chain, integrated solutions demonstrate even greater potential. Collectively, the achievements signify the transition of opto-intelligent agriculture from conceptual exploration to practical implementation. [Conclusions and Prospects] By synergizing optical perception with intelligent decision-making, opto-intelligent agriculture is driving a fundamental transformation in agricultural production. To achieve the transition from merely usable to genuinely effective, a comprehensive advancement framework integrating technology, equipment, talent, and policy must be established. Technologically, efforts should focus on enhancing sensing stability under open-field conditions, developing lightweight and interpretable models, and promoting the domestic development of core components. From a talent perspective, interdisciplinary education and agricultural technology training must be strengthened. From a policy standpoint, improving subsidy mechanisms, digital infrastructure, and innovation-oriented dissemination systems will be essential. Looking forward, through integration with emerging technologies such as 6G communication and digital twin systems, opto-intelligent agriculture is poised to become a cornerstone for ensuring both food security and ecological sustainability.

Key words: opto-intelligent agriculture, data sensing technology, optical sensors, intelligent decision-making systems, multi-source data fusion