Abstract: 【Objective】 This paper aims to review and summarize the research status and development direction of automatic target spraying technology in facility vegetables at home and abroad in the hope of providing theoretical and scientific basis for the development and application of the technology in automatic target spraying robot of vegetable in facility vegetables. 【Methods】 Relevant literatures at home and abroad were collected, field research conducted, and after that, they were systematically summarized and sorted out. 【Results】 Overseas navigation technology mainly adopts path recognition and intelligent obstacle avoidance technology based on GPS, machine vision, ladar and other technologies, and has been widely applied. In China, it mainly adopts electromagnetic induction, road edge acquisition and road identification technology based on GPS, ladar and vision technology. At the present stage, image recognition, infrared imaging, hyperspectrum and deep learn-based pest identification technologies are mainly used in foreign countries, which are relatively mature. At present, image recognition technology is mainly used in China to identify crops with color, texture and shape features. In the developed countries, machine vision, laser active vision and ultrasonic technology combined with sensor identification are employed to identify target crops, and variable speed spraying technology is applied in crops, such as lettuce, tomato and other crops, domestic scholars have developed greenhouse autonomous spraying robot, using machine vision technology to obtain target pest and disease location information, the nozzle is controlled separately to achieve the effect of accurate target application. 【Conclusion】 Navigation technology, pest and disease identification technology and target spraying technology are the core of automatic target spraying technology. Navigation in the greenhouse by using machine vision and laser radar technology compared with the GPS technology is more reliable, flexible, and accurate. In the aspect of pest and disease identification, we should develop pest and disease identification technology combined with hyperspectral and deep learning to improve the efficiency. The identification of target crops and the judgment of dense degree of branches and leaves need more in-depth study.

Key words: facility vegetables; greenhouse; navigation technology; identification of pests and diseases; target spraying; trend

摘要： 【目的】回顾与总结国内外设施蔬菜自动对靶喷药技术的研究现状与进展,为该技术在设施蔬菜自动对靶喷药机器人的发展应用上提供理论和科学依据。【方法】采用相关文献资料、实地调研的方法,汇总、整理及分析。【结果】导航技术国外主要采用基于GPS、机器视觉、激光雷达等技术开发的路径识别及智能避障技术,国内主要采用电磁诱导、基于GPS、激光雷达和视觉技术的道路边缘获取与道路识别技术。病虫害检测现阶段国外主要采用图像识别、红外成像和高光谱及基于深度学习的病虫害识别技术,技术较为成熟,国内现阶段主要采用图像识别技术,利用作物颜色、纹理及形状特征进行识别。国外对靶喷药采用机器视觉、激光主动视觉和超声波技术并结合传感器对目标作物进行识别,利用变速喷药技术在生菜、番茄等作物上进行了应用,国内开发了温室自主喷药机器人,采用机器视觉技术获取靶标病虫害位置信息,对喷头进行单独控制,以达到精准对靶施药的效果。【结论】导航技术、病虫害识别技术及对靶喷药技术是自动对靶喷药技术的核心。导航方面在温室中利用机器视觉和激光雷达技术相比GPS技术更加可靠、灵活,精准度更高,高光谱与病虫害识别技术可提高病虫害识别的效率,对靶喷药技术中目标作物的识别与冠层稠密程度的判断是发展趋势。

关键词: 设施蔬菜, 温室, 导航技术, 病虫害识别, 对靶喷药, 趋势