郑加强，徐幼林，张慧春，周宏平，李秋洁.国内外林业有害生物防治靶标识别技术的研究与展望[J]．林业科学,2023，59（3）：152-166. doi：10.11707/j.1001-7488.LYKX20210996

Zheng Jiaqiang，Xu Youlin，Zhang Huichun，Zhou Hongping，Li Qiujie. Advances and Prospects of Target Recognition Techniques for Forest Pest Control at Home and Abroad[J]．SCIENTIA SILVAE SINICAE, 2023, 59(3):152-166. doi：10.11707/j.1001-7488.LYKX20210996

国内外林业有害生物防治靶标识别技术的研究与展望

摘  要

作为陆地森林生态系统骨架的树木长期遭受有害生物的胁迫，危害林木正常生长，甚至造成林木死亡，对林业可持续发展带来严重威胁，为了实现对林业有害生物靶标及灾害区域的快速、准确检测及监测，需要识别获取林业有害生物防治靶标对象的特征。本文重点围绕生物胁迫之植物病害、植物虫害、鼠（兔）危害、有害植物等直接靶标及其危害的间接靶标的主要特征分析，综述国内外采用不同途径检测和识别林业有害生物防治靶标的技术与装备研究概况。然后建议组织跨学科力量，探索研究基于植物胁迫预警靶标识别、传感网络协同防御控制的林业有害生物防治靶标识别技术及综合防治智能化系统。论文展望了林业有害生物防治靶标识别被动传感技术和主动传感技术的未来发展，提出根据植物在生物胁迫下的物理化学特征表现开展机器视觉、气敏、力敏、热敏、声敏等防治靶标的被动传感技术研究，利用激光雷达扫描、光声效应主动引诱以及电子昆虫和气味主动引诱等主动传感技术，以解决被动传感难以识别特定有害生物靶标的问题，以及研发设计生物、磁敏、湿敏、热敏等新型传感器和研究超声波、LiDAR、辐射、卫星图像等多传感器融合技术，提高林业有害生物靶标的识别精度。最后论文围绕植物根系、野生哺乳动物、土栖白蚁、有害植物等特定靶标提出相应的靶标感知识别技术研究建议。

关键词：林业有害生物；植物保护；靶标识别技术；发展建议

中图分类号：S763；S768；S224.3    文献标识码：A     文章编号：1001-7488（2023）03−0152−15

 Advances and Prospects of Target Recognition Techniques for Forest Pest Control 

at Home and Abroad

Abstract

The trees, as the skeleton of the terrestrial forest ecosystem, suffer from the stresses of forest pests for a long time, which endangers the normal growth of trees, even causes the death of trees, and poses a serious threat to the sustainable development of forestry. It is necessary to recognize and obtain the features of forest pests for realizing the rapid and accurate targets detection and monitoring of forest pests and pest damaged areas. This paper focuses on the analyses of the main features of direct targets of the biological stresses, such as plant diseases, plant insects, rats and rabbits, harmful plants etc. and indirect targets, and summarizes and reviews the research status of techniques and equipments for recognizing and identifying targets for forest pests control through different ways at home and abroad. It is recommended to establish interdisciplinary forces to constantly explore and study the target recognition technology and equipment for forest pest control and integrated control intelligent systems based on the early warning with the target recognition of the plant stresses and the cooperative defense control with the sensor network. This paper looks forward to the future development of passive/active sensing technologies focusing on target identification of forest pests control, and proposes to conduct research on the passive sensing technologies of control targets, including the machine vision, the odor-sensitive, the force-sensitive, the thermal-sensitive and the acoustic-sensitive according to the physical and chemical characteristics of forest plants under biological stresses. The active sensing technologies, such as LiDAR scanning, the photoacoustic effect active attraction and the electronic insects and odoriferous active attraction, would be applied to sense some specific pest targets which the passive sensing technologies are difficult to sense. It is also proposed that the design and development of new sensors such as the biology, the magnetic-sensitive, the humidity-sensitive and the thermal-sensitive sensors, and the multi-sensor fusion technologies for ultrasound, LiDAR, radiation and satellite images, so as to improve the accuracy of forest pest targets recognition. Finally, the paper proposes corresponding research recommendations for target perception and recognition technology around specific targets such as plant roots, wild mammals, terrestrial termites, and harmful plants.

Key words: forest pests; plant protection; targets recognition technology; research prospects