摘要
在虫情监测和害虫防范治理过程中,准确识别害虫是有效解决农业领域虫害问题的重要前提。依靠专家知识和人工经验进行虫情诊断的方式效率较为低下,自动化和智能化水平较差,而采用深度学习、计算机视觉等智能化技术手段可以大幅度提升害虫识别过程的效率、准确度,并降低人工成本。概述了基于深度学习的害虫识别技术发展现状,分析深度学习技术在害虫图像识别领域的实现原理和优势,阐述国内外专家学者在基于深度学习的害虫识别技术领域的最新研究进展,提出该技术领域面临的挑战,并对发展方向进行预测。该文可为深入开展害虫识别和分类技术在智慧农业上的应用研究提供参考。
In process of pest monitoring and pest control,accurate identification of pests is an important prerequisite for effectively solving pest problems in agricultural field.Relying on expert knowledge and planting experience to diagnose insect pests on crops is inefficient and the level of automation and intelligence is poor.The use of intelligent technologies such as deep learning and computer vision can greatly improve efficiency and accuracy of pest identification process and reduce labor costs.Development status of pest recognition technology based on deep learning was outlined.Implementation principles and advantages of deep learning technology in the field of pest image recognition were analyzed.Latest research progress of domestic and foreign experts and scholars in the field of pest recognition technology based on deep learning were described.Finally,challenges facing this field were summarized and future development direction was forecasted.It could provide an important theoretical reference for further research of pest identification and classification technology in smart agriculture.
作者
庞海通
蔡卫明
马龙华
苏宏业
PANG Haitong;CAI Weiming;MA Longhua;SU Hongye(NingboTech University,Ningbo Zhejiang 315100,China;State Key Laboratory of Industrial Control Technology,Zhejiang University,Hangzhou Zhejiang 310027,China)
出处
《农业工程》
2020年第10期19-24,共6页
AGRICULTURAL ENGINEERING
基金
国家自然科学基金项目(项目编号:31702393)
宁波市公益技术计划重点类项目(项目编号:2019C10098)
浙江大学工业控制技术国家重点实验室开放课题(项目编号:ICT20007)。
关键词
农业
害虫
识别
深度学习
agriculture
pests
recognition
deep learning