Authors

Abstract

The identification of farmland weeds and variable rate herbicide spraying technology are core components of precision agriculture, playing a significant role in enhancing agricultural productivity, reducing pesticide usage, and protecting the ecological environment. Currently, global agriculture faces dual challenges of increasing resource constraints and rising environmental protection demands. This technology, by precisely locating weed distribution and adjusting pesticide application rates accordingly, has become a key approach to breaking the vicious cycle of "pesticide overuse-weed resistance-ecological pollution." Based on bibliometric methods and using the Web of Science database as the data source, this study retrieved literature related to farmland weed identification and variable rate herbicide spraying from 2005 to 2024. VOSviewer software was employed for visual analysis, systematically examining the temporal evolution characteristics, regional collaboration networks, institutional contributions, and keyword clustering patterns in this field. The results indicate that research in this area entered a rapid development phase after 2018, driven significantly by artificial intelligence technology. Research hotspots focus on image recognition algorithms, multi-source data fusion, variable rate herbicide spraying system design, and field application validation. Current studies face challenges in adaptability to complex environments and multi-scale data coordination. Future efforts should strengthen lightweight recognition model optimization, space-air-ground integrated data fusion, cost-effective smart equipment development, and interdisciplinary collaboration to provide technical support for the sustainable development of precision agriculture.

Abstract in Chinese

农田杂草识别与变量施药技术是精准农业的核心组成部分，对提升农业生产效率、降低农药使用量及保护生态环境具有重要意义。当前，全球农业面临资源约束加剧、生态环保需求提升的双重挑战，该技术通过精准定位杂草分布并按需调控施药量，已成为破解“农药滥用-杂草抗药性-生态污染”恶性循环的关键途径。本研究基于文献计量学方法，以Web of Science数据库为数据源，检索2005-2024年农田杂草识别及变量施药领域相关文献，运用VOSviewer软件进行可视化分析，系统梳理该领域研究的时间演化特征、地域合作网络、机构贡献及关键词聚类规律。结果表明：该领域研究在2018年后进入快速发展期，受人工智能技术驱动显著；研究热点集中于图像识别算法、多源数据融合、变量施药系统设计及田间应用验证等方面。当前研究在复杂环境适应性、多尺度数据协同等方面存在挑战，未来需加强轻量级识别模型优化、空天地一体化数据融合、低成本智能装备研发及跨学科协同，为精准农业可持续发展提供技术支撑。