昆虫检测识别技术研究进展 ——形态鉴定、分子检测与多模态智能识别
Advances in insect detection and identification technologies —morphological identification, molecular detection, and multimodal intelligent recognition
张 柳1** 葛 阳2 陈 克1 李红卫1 于艳雪1***
点击：52次 下载：4次
DOI：10.7679/j.issn.20951353.2026.106
作者单位：1. 中国质量检验检测科学研究院植物安全与质量研究所, 北京 100176； 2. 中国中医科学院中药资源中心，道地药材品质保障与资源持续利用全国重点实验室，北京 100700
中文关键词：昆虫；图像识别；分子检测；仿生嗅觉检测技术；声学检测技术；人工智能
英文关键词：insects; image recognition; molecular identification; biomimetic olfactory detection technology; acoustic detection technology; artificial intelligence
中文摘要：精准的物种鉴定是害虫监测防治及生物资源开发利用的基础，对害虫的早期预警和精准防控尤为关键。本文系统综述了目前昆虫检测识别技术的研究进展，从经典形态鉴定、分子生物学检测，到基于图像、气味与声学检测等多模态信息智能识别的范式演进。传统形态鉴定构成整个技术链的基石，为后续基因、图像、化学和声音信号的特异性采集提供分类学基准，精准、便携、快速的检测鉴定方法逐渐成为现今的研发重点。在实际应用中，单一模态检测技术的固有局限性难以满足高效防控需求，催生多技术融合路径。未来研究应聚焦多模态标准化数据库建设、便携式现场装备研发及融合算法的可解释性提升，以期构建高效、精准、可部署的智能识别系统，为延缓重大害虫传播扩散与精准防控提供技术支撑。
英文摘要：Accurate species identification serves as the foundation for pest monitoring and management, as well as for the development and utilization of biological resources. It is particularly critical for the early warning and precise control of insect pests. This review systematically summarizes current progress in research on insect detection and identification technology, from classical morphological identification and molecular biological detection, to the paradigm evolution of multimodal intelligent recognition based on image, odor and acoustic information. Traditional morphological identification constitutes the cornerstone of the entire technical chain, providing a taxonomic benchmark for the specific collection of subsequent genetic, image, chemical and acoustic signals. Accurate, portable, and rapid, detection and identification methods have become a key focus of research and development. In practical applications, the inherent limitations of single-modal detection techniques make it difficult to meet the demands of efficient prevention and control, thus giving rise to the path of multi technology integration. Future research should focus on establishing multimodal standardized databases, developing portable field equipment, and improving the interpretability of fusion algorithms. These advances are expected to produce efficient, accurate, and deployable, intelligent identification systems, thereby providing essential technical support for curbing the spread of major pests and improving the precision of management strategies.