技术领域technical field
本发明涉及一种病虫害预警系统，具体涉及一种基于机器视觉的病虫害预警系统和方法。The invention relates to a disease and insect pest early warning system, in particular to a machine vision-based disease and insect pest early warning system and method.
背景技术Background technique
在工业化、信息化、城镇化深入发展中同步推进农业现代化，是“十二五”时期的一项重大任务。“十二五”是全面建设小康社会的关键时期，是深化改革开放、加快转变经济发展方式的攻坚时期，是加快发展现代农业的重要机遇期。Simultaneously promoting agricultural modernization while industrialization, informatization, and urbanization are in-depth development is a major task during the "Twelfth Five-Year Plan" period. The "Twelfth Five-Year Plan" is a critical period for building a well-off society in an all-round way, a critical period for deepening reform and opening up, accelerating the transformation of economic development methods, and an important period of opportunity for accelerating the development of modern agriculture.
数字化农业和精准化作业是现代农业发展的方向和要求。在作物病情虫情分析方面，数字农业要求快速、准确地获取植物受病虫害侵染的信息，从而指导植物生长过程中的精细化管理。Digital agriculture and precision operations are the direction and requirements of modern agricultural development. In terms of crop disease and insect pest analysis, digital agriculture requires fast and accurate acquisition of plant disease and insect pest infestation information, so as to guide the fine management of plant growth.
国际上，日本在20世纪末已经在技术密集型的设施园艺领域开发了多种病虫害预警系统，不但节省了人力物力，还大大提高了病虫害防治效果；荷兰农业环境工程研究所开发的病虫害预警的专家系统，运用图像处理技术和专家系统技术结合，在应用方面取得了良好效果；而美国、英国等国家的温室大棚在采用智能控制系统的同时结合机器视觉技术对病虫害的等级程度、范围进行了预警。Internationally, Japan has developed a variety of early warning systems for plant diseases and insect pests in the field of technology-intensive facility gardening at the end of the 20th century, which not only saves manpower and material resources, but also greatly improves the effect of disease and insect pest control; The system, using the combination of image processing technology and expert system technology, has achieved good results in application; while the greenhouses in the United States, the United Kingdom and other countries use intelligent control systems and combine machine vision technology to give early warning of the level and scope of pests and diseases. .
目前农作物病虫害预警的主要难点在于要完成复杂背景下的病斑分割以及采用一种识别率较高但计算量较小的识别算法。但目前先进的温室计算机监测预警系统仍依赖国外进口技术，核心技术仍然掌握于国外。因此在对温室环境中农作物病虫害预警方面的研究仍需要进一步的努力和创新。At present, the main difficulty in the early warning of crop diseases and insect pests is to complete the segmentation of diseased spots in complex backgrounds and to adopt a recognition algorithm with a high recognition rate but a small amount of calculation. However, the current advanced greenhouse computer monitoring and early warning system still relies on imported technology, and the core technology is still mastered abroad. Therefore, further efforts and innovations are still needed in the research on the early warning of crop diseases and insect pests in the greenhouse environment.
发明内容Contents of the invention
本发明的目的是为了解决现有技术中对农作物病虫害实现预警的缺陷，提供一种基于机器视觉的病虫害预警系统和方法，由改进的otsu分割算法及KNN智能分类算法为核心解决上述问题。The purpose of the present invention is to solve the defects of early warning of crop diseases and insect pests in the prior art, and provide a system and method for early warning of plant diseases and insect pests based on machine vision. The improved otsu segmentation algorithm and KNN intelligent classification algorithm are used as the core to solve the above problems.
为了实现上述目的，本发明的构思是：In order to achieve the above object, design of the present invention is:
基于机器视觉的病虫害预警系统由软件和硬件两个部分共同组成，根据功能不同又可将该系统划分图像与环境参数采集系统以及图像处理系统；其中，图像与环境参数采集系统由现场佳能EOS7D类型相机及多种环境传感器获取图像监The early warning system of diseases and insect pests based on machine vision is composed of software and hardware. According to different functions, the system can be divided into image and environmental parameter acquisition system and image processing system; among them, the image and environmental parameter acquisition system is composed of Canon EOS7D type on site. Cameras and various environmental sensors acquire image monitoring
测数据和环境参数，再将数据传输致图像处理系统进行相应的处理；图像处理系统由警源预警和警兆预警及信息处理组成。警源预警是通过将实时的现场环境参数与根据专家提供的适宜农作物生长的环境参数进行比对，进行病虫害的初步预警；警兆预警是利用机器视觉图像处理技术对现场传感器采集的图像数据进行处理，判断病虫害的类别；信息处理实现对系统的处理结果以及相关记录写入数据库，同时提供查询功能。每个部分各有其任务，各个模块配套软件来整合成一个有机的整体。The measured data and environmental parameters are then transmitted to the image processing system for corresponding processing; the image processing system is composed of alarm source early warning, warning sign early warning and information processing. Alarm source early warning is to carry out preliminary early warning of diseases and insect pests by comparing real-time on-site environmental parameters with environmental parameters suitable for crop growth provided by experts; warning sign early warning is to use machine vision image processing technology to analyze image data collected by on-site sensors. Processing, judging the type of pests and diseases; information processing realizes the processing results and related records of the system are written into the database, and at the same time provides query functions. Each part has its own task, each module supporting software to integrate into an organic whole.
所述的图像与环境参数采集系统由佳能EOS7D相机及环境参数传感器（主要测量参数温度、湿度、光强）组成；图像与数据采集系统测量的数据通过无线传输的方式传到图像处理系统。The image and environmental parameter acquisition system is composed of a Canon EOS7D camera and an environmental parameter sensor (mainly measuring parameters temperature, humidity, and light intensity); the data measured by the image and data acquisition system is transmitted to the image processing system through wireless transmission.
所述的佳能EOS7D相机主要参数指标为：镜头型号：EF-S 18-135mm f/3.5-5.6IS；实际焦距：f=18-135mm；传感器尺寸APS画幅（22.3*14.9mm）；最高分辨率5184×3456；影像处理器DIGIC 4+DIGIC 4；取景器：（类型：眼平五棱镜视野率：垂直/水平方向约100%；放大倍率：约1.0倍（-1m，使用50mm镜头对无限远处对焦）；眼点：约22mm（自目镜透镜中央起-1m）内置屈光度调节：-3.0-+1.0m（dpt）；对焦屏：固定式；构图辅助：网格线和电子水准仪；反光镜：快回型景深预视）。The main parameters of the Canon EOS7D camera mentioned are: lens model: EF-S 18-135mm f/3.5-5.6IS; actual focal length: f=18-135mm; sensor size APS format (22.3*14.9mm); highest resolution 5184×3456; image processor DIGIC 4+DIGIC 4; viewfinder: (type: eye-level pentaprism field of view: about 100% in vertical/horizontal directions; magnification: about 1.0 times (-1m, using a 50mm lens for infinity eye point: about 22mm (-1m from the center of the eyepiece lens), built-in diopter adjustment: -3.0-+1.0m (dpt); focusing screen: fixed; composition aids: grid lines and electronic level; mirror : Quick-return type depth-of-field preview).
所述的图像处理系统由警源预警和警兆预警及信息处理组成；先由警源预警进行初步预警，再由警兆预警对病虫害进行识别与分类。同时配备供查询历史记录的信息处理功能。The image processing system is composed of alarm source early warning, warning sign early warning and information processing; the alarm source early warning performs preliminary early warning, and then the warning sign early warning identifies and classifies pests and diseases. At the same time, it is equipped with an information processing function for querying historical records.
所述的警源预警是由现场多种环境传感器获取环境参数 （温度℃、湿度%、光强Lux），将实时环境参数与根据专家提供的适宜农作物生长的环境参数进行比对，及时的给出预警信息，在还没有发病之前就可以预防病虫害的发生。The alarm source early warning is to obtain environmental parameters (temperature °C, humidity %, light intensity Lux) by various environmental sensors on site, compare the real-time environmental parameters with the environmental parameters suitable for the growth of crops provided by experts, and timely give Early warning information can prevent the occurrence of pests and diseases before the onset of disease.
所述的警兆预警是由图像滤波、图像分割、特征提取、形态学操作、图像锐化、训练自学习模块组成。所述的警兆预警是整个系统的核心，实现对病虫害的分类识别，直接影响系统的预警效果。The warning sign and early warning is composed of image filtering, image segmentation, feature extraction, morphological operation, image sharpening, training and self-learning modules. The warning sign and early warning is the core of the whole system, which realizes the classification and identification of pests and diseases, and directly affects the early warning effect of the system.
所述的图像滤波主要有中值滤波、均值滤波、KNN平滑滤波、阈值滤波等滤波方式。从而实现对图像在采集、量化、传输的过程中产生的噪声进行滤波，以降低噪声对后续图像处理和识别的影响。The image filtering mainly includes filtering methods such as median filtering, mean filtering, KNN smoothing filtering, and threshold filtering. In this way, the noise generated in the process of image acquisition, quantification, and transmission can be filtered to reduce the impact of noise on subsequent image processing and recognition.
所述的图像分割主要有OTSU自动阈值分割、色值分割、自动阈值分割、一维最大熵法分割等分割方式。实现对经预处理后的图像进行分割，分割出完整病斑，为后续提取特征做准备，图像分割是否完整将直接影响提取特征是否充分。The image segmentation mainly includes OTSU automatic threshold segmentation, color value segmentation, automatic threshold segmentation, one-dimensional maximum entropy segmentation and other segmentation methods. Realize the segmentation of the preprocessed image, segment the complete lesion, and prepare for the subsequent feature extraction. Whether the image segmentation is complete will directly affect whether the extracted feature is sufficient.
所述的特征提取主要提取病斑形态学特征（周长L、面积S、圆度C、复杂度E）和颜色特征（色调H、色饱和度S、亮度 I），这些特征很好的表达了每种病斑的本质特征，得到的特征参数是判别病斑种类的重要依据。The feature extraction described mainly extracts lesion morphological features (perimeter L, area S, roundness C, complexity E) and color features (hue H, color saturation S, brightness I), which are well expressed The essential characteristics of each lesion were obtained, and the obtained characteristic parameters were an important basis for distinguishing the types of lesion.
所述的图像锐化主要有Laplace算子、prewitt算子、梯度算子、kirsch算子、sobel算子。为了利于图像分割可采取不同的锐化方式，使分割达到最好的效果。The image sharpening mainly includes Laplace operator, prewitt operator, gradient operator, kirsch operator, and sobel operator. In order to facilitate image segmentation, different sharpening methods can be adopted to achieve the best segmentation effect.
所述的形态学操作主要手段有开操作和闭操作。形态学操作主要是针对分割完的图像，根据分割完的效果采取适当的形态学操作，以便于提取特征。The main means of the morphological operation include opening operation and closing operation. The morphological operation is mainly for the segmented image, and the appropriate morphological operation is taken according to the segmented effect in order to extract features.
所述的训练自学模块主要包含训练和识别两个模块。训练模块采取的有监督训练，即在识别前拿大量的已知类别的图像数据输入系统，来建立特征库。识别的过程即输入待识别图像与已知的特征的类库根据智能识别算法来判别图像种类。The training self-study module mainly includes two modules of training and recognition. The supervised training adopted by the training module is to input a large amount of image data of known categories into the system before recognition to establish a feature library. The process of recognition is to input the image to be recognized and the class library of known features to distinguish the image type according to the intelligent recognition algorithm.
所述的信息处理功能使软件直接与数据库相连接（SQL serve 2008），实现对系统的处理结果以及相关记录写入数据库，同时提供查询功能，使得用户可以实时查询病虫害历史处理记录，了解发病历程。The information processing function makes the software directly connected to the database (SQL serve 2008), realizes the processing results of the system and related records are written into the database, and provides a query function at the same time, so that users can query the historical processing records of diseases and insect pests in real time, and understand the disease process .
根据上述发明构思，本发明采用下述技术方案：According to above-mentioned inventive concept, the present invention adopts following technical scheme:
一种基于机器视觉的病虫害预警系统和方法，其特征在于，包含佳能EOS7D相机（1）安装于温室大棚角落固定位置，用于温室内作物的图像数据采集、环境参数集成传感器（2）安装于温室大棚顶部，用于温室室内多种环境参数数据采集、图像处理系统（3）位于温室中央控制室内，用于对温室内采集的图像数据和环境参数数据进行处理、记录。佳能EOS7D相机（1）和环境参数集成传感器（2）连接图像处理系统；A machine vision-based pest and disease early warning system and method, characterized in that it includes a Canon EOS7D camera (1) installed at a fixed position in the corner of a greenhouse for image data collection of crops in the greenhouse, and an integrated environmental parameter sensor (2) installed in The top of the greenhouse is used for data collection of various environmental parameters in the greenhouse, and the image processing system (3) is located in the central control room of the greenhouse, which is used to process and record the image data and environmental parameter data collected in the greenhouse. Canon EOS7D camera (1) and environmental parameter integrated sensor (2) are connected to the image processing system;
所述佳能EOS7D相机（1）主要参数指标为：有效像素1800万；实际焦距：f=18-135mm；文件格式：JPEG，RAW(14位)，可以同时记录RAW+JPEG。The main parameters of the Canon EOS7D camera (1) are: 18 million effective pixels; actual focal length: f=18-135mm; file format: JPEG, RAW (14 bits), and can record RAW+JPEG at the same time.
所述环境参数集成传感器（2）由温度、湿度、光强多种环境参数传感器集合而成；用于测量温室内的各种环境参数，利用无线网络传输至图像处理系统（3），用于警源预警的依据。The environmental parameter integrated sensor (2) is composed of various environmental parameter sensors such as temperature, humidity, and light intensity; it is used to measure various environmental parameters in the greenhouse, and is transmitted to the image processing system (3) using a wireless network for Basis for alarm source warning.
所述图像处理系统（3）具体包括：The image processing system (3) specifically includes:
警源预警模块（3-1），根据现场多种环境传感器获取环境参数 ，实时环境参数与农作物适宜生长环境参数进行匹配给出初步预警，并在软件界面给警示信息。The alarm source early warning module (3-1) acquires environmental parameters according to various environmental sensors on site, and matches the real-time environmental parameters with the suitable growth environment parameters of crops to give a preliminary early warning, and gives warning information on the software interface.
警兆预警模块（3-2），在警源预警的基础上，根据现场佳能EOS7D相机获取图像数据 ，利用机器视觉中图像处理技术对图像数据进行处理。通过预处理、特征提取、对病虫害种类进行识别分类；The warning sign and early warning module (3-2), based on the early warning of the warning source, obtains image data according to the on-site Canon EOS7D camera, and uses the image processing technology in machine vision to process the image data. Identify and classify the types of pests and diseases through preprocessing and feature extraction;
信息处理模块（3-3），对系统的处理结果以及相关记录写入数据库，同时也提供查询功能，使得用户实时查询病虫害历史处理记录，了解发病历程。The information processing module (3-3) writes the processing results of the system and related records into the database, and also provides a query function, allowing users to query the historical processing records of diseases and insect pests in real time and understand the disease history.
一种基于机器视觉的病虫害预警系统和方法，采用上述系统进行操作，其特征在于操作步骤如下：A machine vision-based early warning system and method for pests and diseases, using the above-mentioned system for operation, is characterized in that the operation steps are as follows:
步骤1：佳能EOS7D相机（1）与环境参数集成传感器（2）每隔3小时采集 一次数据，并通过无线网络将数据传输至图像处理系统（3）。Step 1: Canon EOS7D camera (1) and environmental parameter integrated sensor (2) collect data every 3 hours, and transmit the data to the image processing system (3) through wireless network.
步骤2：图像处理系统（3）中，警源预警模块（3-1）启动，将现场采集的环境参数数据与预先设定好的专家经验数值进行匹配，判断是否在合理范围内。Step 2: In the image processing system (3), the alarm source warning module (3-1) is started, and the environmental parameter data collected on the spot is matched with the preset expert experience value to judge whether it is within a reasonable range.
步骤3：若环境参数数据不在合理范围内，则启动警兆预警模块（3-2）。针对佳能EOS7D相机（1）采集的图像数据进行图像处理，进一步判别 是否有病虫害，以及病虫害的类别。并及时给出预警信息。Step 3: If the environmental parameter data is not within a reasonable range, start the warning sign and early warning module (3-2). Image processing is performed on the image data collected by the Canon EOS7D camera (1) to further identify whether there are pests and diseases, and the types of pests and diseases. And give early warning information in time.
步骤4：信息处理模块（3-3），对于警源预警模块（3-1）及警兆预警模块 （3-2）的处理数据及处理结果写入数据库待查。Step 4: The information processing module (3-3) writes the processing data and processing results of the alarm source early warning module (3-1) and warning sign early warning module (3-2) into the database for investigation.
本发明与现有技术相比，具有如下显而易见的突出实质性特点和显著优点：Compared with the prior art, the present invention has the following obvious outstanding substantive features and significant advantages:
本发明创新性的将机器视觉技术应用与农作物的病虫害预警，不仅在现场使用先进的传感器代替人工检测而且实现了从环境变化到病兆两个维度的病虫害智能化预警，并在分割算法及智能识别算法上做出优化，提高了预警的及时性与准确性，大大更新了管理效率的模式。The invention innovatively applies the machine vision technology to the early warning of crop diseases and insect pests, not only using advanced sensors on site instead of manual detection, but also realizing the intelligent early warning of diseases and insect pests in two dimensions from environmental changes to disease symptoms, and in the segmentation algorithm and intelligent The optimization of the recognition algorithm has improved the timeliness and accuracy of early warning, and greatly updated the mode of management efficiency.
附图说明Description of drawings
图1为病虫预警系统结构示意图Figure 1 is a schematic diagram of the structure of the disease and pest early warning system
图2图像与环境参数采集系统Figure 2 Image and environmental parameter acquisition system
图3为警源预警运行界面图Figure 3 is the operation interface diagram of alarm source early warning
图4为警兆预警运行界面图Figure 4 is the operation interface diagram of warning sign and early warning
图5为图像滤波、分割、特征提取界面图Figure 5 is an interface diagram of image filtering, segmentation, and feature extraction
图6为图像识别预警界面图Figure 6 is the image recognition early warning interface diagram
图7为系统信息处理界面图。Figure 7 is a system information processing interface diagram.
具体实施方式detailed description
为使对本发明的预警结构特征及所达成的功效有更进一步的了解与认识，用以优选的实施例及附图配合详细的说明，说明如下：In order to have a further understanding and understanding of the early warning structural features and the achieved effects of the present invention, the preferred embodiments and accompanying drawings are used for a detailed description, as follows:
实施例一：Embodiment one:
参观图1~~图7，本基于机器视觉的病虫害预警系统和方法，其特征在于，包含佳能EOS7D相机（1）安装于温室大棚角落固定位置，用于温室内作物的图像数据采集、环境参数集成传感器（2）安装于温室大棚顶部，用于温室室内多种环境参数数据采集、图像处理系统（3）位于温室中央控制室内，用于对温室内采集的图像数据和环境参数数据进行处理、记录。佳能EOS7D相机（1）和环境参数集成传感器（2）连接图像处理系统；Visit Figure 1~~Figure 7, this machine vision-based pest and disease early warning system and method is characterized in that it includes a Canon EOS7D camera (1) installed at a fixed position in the corner of the greenhouse for image data collection and environmental parameters of crops in the greenhouse The integrated sensor (2) is installed on the top of the greenhouse for data collection of various environmental parameters in the greenhouse. The image processing system (3) is located in the central control room of the greenhouse and is used to process the image data and environmental parameter data collected in the greenhouse. Record. Canon EOS7D camera (1) and environmental parameter integrated sensor (2) are connected to the image processing system;
实施例二：本实施例与实施例一基本相同，特别之处如下：Embodiment 2: This embodiment is basically the same as Embodiment 1, and the special features are as follows:
所述佳能EOS7D相机（1）主要参数指标为：有效像素1800万；实际焦距：f=18-135mm；文件格式：JPEG，RAW(14位)，可以同时记录RAW+JPEG。The main parameters of the Canon EOS7D camera (1) are: 18 million effective pixels; actual focal length: f=18-135mm; file format: JPEG, RAW (14 bits), and can record RAW+JPEG at the same time.
所述环境参数集成传感器（2）由温度、湿度、光强多种环境参数传感器集合而成；用于测量温室内的各种环境参数，利用无线网络传输至图像处理系统（3），用于警源预警的依据。The environmental parameter integrated sensor (2) is composed of various environmental parameter sensors such as temperature, humidity, and light intensity; it is used to measure various environmental parameters in the greenhouse, and is transmitted to the image processing system (3) using a wireless network for Basis for alarm source warning.
所述图像处理系统（3）具体包括：The image processing system (3) specifically includes:
警源预警模块（3-1），根据现场多种环境传感器获取环境参数 ，实时环境参数与农作物适宜生长环境参数进行匹配给出初步预警，并在软件界面给警示信息。The alarm source early warning module (3-1) acquires environmental parameters according to various environmental sensors on site, and matches the real-time environmental parameters with the suitable growth environment parameters of crops to give a preliminary early warning, and gives warning information on the software interface.
警兆预警模块（3-2），在警源预警的基础上，根据现场佳能EOS7D相机获取图像数据 ，利用机器视觉中图像处理技术对图像数据进行处理。通过预处理、特征提取、对病虫害种类进行识别分类；The warning sign and early warning module (3-2), based on the early warning of the warning source, obtains image data according to the on-site Canon EOS7D camera, and uses the image processing technology in machine vision to process the image data. Identify and classify the types of pests and diseases through preprocessing and feature extraction;
信息处理模块（3-3），对系统的处理结果以及相关记录写入数据库，同时也提供查询功能，使得用户实时查询病虫害历史处理记录，了解发病历程。The information processing module (3-3) writes the processing results of the system and related records into the database, and also provides a query function, allowing users to query the historical processing records of diseases and insect pests in real time and understand the disease history.
实施例三：Embodiment three:
本基于机器视觉的病虫害预警系统和方法，采用上述系统进行操作，其特征在于操作步骤如下：This machine vision-based early warning system and method for pests and diseases uses the above-mentioned system to operate, and is characterized in that the operation steps are as follows:
步骤1：佳能EOS7D相机（1）与环境参数集成传感器（2）每隔3小时采集 一次数据，并通过无线网络将数据传输至图像处理系统（3）。Step 1: Canon EOS7D camera (1) and environmental parameter integrated sensor (2) collect data every 3 hours, and transmit the data to the image processing system (3) through wireless network.
步骤2：图像处理系统（3）中，警源预警模块（3-1）启动，将现场采集的环境参数数据与预先设定好的专家经验数值进行匹配，判断是否在合理范围内。Step 2: In the image processing system (3), the alarm source warning module (3-1) is started, and the environmental parameter data collected on the spot is matched with the preset expert experience value to judge whether it is within a reasonable range.
步骤3：若环境参数数据不在合理范围内，则启动警兆预警模块（3-2）。针 对佳能EOS7D相机（1）采集的图像数据进行图像处理，进一步判别是否有病虫害，以及病虫害的类别。并及时给出预警信息。Step 3: If the environmental parameter data is not within a reasonable range, start the warning sign and early warning module (3-2). Image processing is performed on the image data collected by the Canon EOS7D camera (1) to further identify whether there are pests and diseases, and the types of pests and diseases. And give early warning information in time.
步骤4：信息处理模块（3-3），对于警源预警模块（3-1）及警兆预警模块（3-2）的处理数据及处理结果写入数据库待查。Step 4: The information processing module (3-3) writes the processing data and processing results of the alarm source early warning module (3-1) and warning sign early warning module (3-2) into the database for investigation.
图1~~图7所示，本基于机器视觉的病虫害预警系统和方法主要由图像与环境参数采集系统及图像处理系统（主要包含警源预警、警兆预警、信息处理三部分组成），各功能模块相互互补搭配完成整体工作任务。As shown in Figures 1 to 7, the machine vision-based pest and disease early warning system and method are mainly composed of an image and environmental parameter acquisition system and an image processing system (mainly composed of three parts: alarm source early warning, warning sign early warning, and information processing). The functional modules complement each other and cooperate to complete the overall task.
首先当软件运行时，系统调用数据库中现场各传感器传回的现场环境数据参数（温度℃、湿度%、光强Lux），并如图3所示显示在警源预警的界面上，此时程序后台把这些现场的环境参数与根据专家建议的农作物适宜的适宜的环境参数变化范围进行比对，判断此时的环境参数是否在适宜的范围，并给出相应的处理建议，并显示在警源预警的界面上。这样实现对病虫害预警的初步预警，预防在发病之前。First, when the software is running, the system calls the field environmental data parameters (temperature °C, humidity%, light intensity Lux) returned by the field sensors in the database, and displays them on the alarm source warning interface as shown in Figure 3. At this time, the program The background compares these on-site environmental parameters with the appropriate range of environmental parameters recommended by experts for crops to judge whether the environmental parameters at this time are within the appropriate range, and give corresponding processing suggestions, which are displayed in the alarm source warning interface. In this way, the preliminary early warning of the early warning of the disease and insect pest is realized, and the prevention is before the onset of the disease.
在警兆界面上，系统调用现场的图像数据参数显示在界面上（如图4所示）并进行相应的处理。如图5所示，图像经由中值滤波，并用OTSU自动阈值分割算法进行分割，得到分割完的完整病斑。在此基础上，程序利用对图像的全屏扫描同时提取病斑形态学特征（周长L、面积S、圆度C、复杂度E）及颜色特征（色调H、色饱和度S、亮度 I），并把这特征参数实时显示在软件界面上（如图5所示），并由最近邻分类算法(kNN）对病斑进行识别分类，并给出判别结果（如图5所示）。并且这些处理的结果实时的写入数据库中，以便后续的查询与管理。On the warning sign interface, the image data parameters of the system call site are displayed on the interface (as shown in Figure 4) and corresponding processing is performed. As shown in Figure 5, the image was filtered by the median value and segmented with the OTSU automatic threshold segmentation algorithm to obtain the segmented complete lesion. On this basis, the program simultaneously extracts lesion morphological features (perimeter L, area S, roundness C, complexity E) and color features (hue H, color saturation S, brightness I) by scanning the image in full screen , and the characteristic parameters are displayed on the software interface in real time (as shown in Figure 5), and the nearest neighbor classification algorithm (kNN) is used to identify and classify the lesions, and the discrimination results are given (as shown in Figure 5). And the results of these processes are written into the database in real time for subsequent query and management.
在信息处理的界面上，如图7所示可以实时显示近期系统的处理记录及相关参数，也可以从数据库中查询历史的记录，便于对于病虫害的全程监控以及后续调用相关数据进行研究。On the information processing interface, as shown in Figure 7, recent system processing records and related parameters can be displayed in real time, and historical records can also be queried from the database, which is convenient for the whole monitoring of pests and diseases and the subsequent call of relevant data for research.
在现场传感器的配合下，本系统实现了从警源预警到警兆预警双重病虫害预警，同时配备提供用户可以实时查询预警记录的信息处理功能。先由实时环境参数与农作物适宜生长环境参数进行匹配给出初步预警，然后使用机器视觉中图像处理技术对实时图像数据进行处理（滤波、分割、特征提取、识别分类），实现对番茄早疫病和叶霉病的自动识别与分类，给出目前农作物的病虫害发病情况，为用户提供决策支持及历史信息记录查询。也大大减少了人力物力，同时又提升了温室农作物病虫害管理的效率。With the cooperation of on-site sensors, this system realizes double pest and disease early warning from alarm source early warning to warning sign early warning. At the same time, it is equipped with an information processing function that provides users with real-time query of early warning records. First, the preliminary warning is given by matching the real-time environmental parameters with the suitable growth environment parameters of the crops, and then the real-time image data is processed (filtering, segmentation, feature extraction, identification and classification) using the image processing technology in machine vision to realize the detection of tomato early blight and Automatic identification and classification of leaf mold, giving the current incidence of crop diseases and insect pests, providing decision support and historical information record query for users. It also greatly reduces manpower and material resources, and at the same time improves the efficiency of greenhouse crop pest management.
以上显示和描述了本发明的基本原理、主要特征和本发明的优点。本行业的技术人员应该了解，本发明不受上述实施例的限制，上述实施例和说明书中描述的只是本发明的原理，在不脱离本发明精神和范围的前提下本发明还会有各种变化和改进，这些变化和改进都落入要求保护的本发明的范围内。本发明要求的保护范围由所附的权利要求书及其等同物界定。The basic principles, main features and advantages of the present invention have been shown and described above. Those skilled in the art should understand that the present invention is not limited by the above-mentioned embodiments. What are described in the above-mentioned embodiments and the description are only the principles of the present invention. Variations and improvements, which fall within the scope of the claimed invention. The scope of protection required by the present invention is defined by the appended claims and their equivalents.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610254139.8ACN105850930A (en) | 2016-04-23 | 2016-04-23 | Early warning system and method of diseases and insect pests based on machine vision |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610254139.8ACN105850930A (en) | 2016-04-23 | 2016-04-23 | Early warning system and method of diseases and insect pests based on machine vision |
| Publication Number | Publication Date |
|---|---|
| CN105850930Atrue CN105850930A (en) | 2016-08-17 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610254139.8APendingCN105850930A (en) | 2016-04-23 | 2016-04-23 | Early warning system and method of diseases and insect pests based on machine vision |
| Country | Link |
|---|---|
| CN (1) | CN105850930A (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106779188A (en)* | 2016-11-30 | 2017-05-31 | 深圳前海弘稼科技有限公司 | Plant pest Forecasting Methodology and device in a kind of plantation equipment |
| CN106780443A (en)* | 2016-11-30 | 2017-05-31 | 深圳前海弘稼科技有限公司 | The method for early warning of crop pest and the prior-warning device of crop pest |
| CN106846462A (en)* | 2016-12-30 | 2017-06-13 | 北京农业信息技术研究中心 | Insect identifying device and method based on three-dimensional simulation |
| CN107168367A (en)* | 2017-06-08 | 2017-09-15 | 中科院合肥技术创新工程院 | Accurate pesticide spraying system based on machine vision industry |
| CN107272620A (en)* | 2017-06-23 | 2017-10-20 | 深圳市盛路物联通讯技术有限公司 | A kind of method and device of the intelligent monitoring greenhouse based on Internet of Things |
| CN107301605A (en)* | 2017-06-29 | 2017-10-27 | 深圳前海弘稼科技有限公司 | The control method and device of crop pest |
| CN107357271A (en)* | 2017-06-30 | 2017-11-17 | 深圳前海弘稼科技有限公司 | The prevention and controls of chamber crop pest and disease damage, preventing and treating system |
| CN108040997A (en)* | 2017-12-25 | 2018-05-18 | 仲恺农业工程学院 | Insect pest monitoring method based on machine vision |
| WO2019126971A1 (en)* | 2017-12-25 | 2019-07-04 | 仲恺农业工程学院 | Pest disaster monitoring method based on machine vision |
| CN110167344A (en)* | 2016-11-11 | 2019-08-23 | 生物生态服务提供商 | System and method for detecting flying animals |
| CN110427922A (en)* | 2019-09-03 | 2019-11-08 | 陈�峰 | One kind is based on machine vision and convolutional neural networks pest and disease damage identifying system and method |
| IT201900002249A1 (en)* | 2019-02-15 | 2020-08-15 | Agrorobotica S R L | System and method of predicting the risk of insect attack |
| CN112034912A (en)* | 2020-09-03 | 2020-12-04 | 山东锋士信息技术有限公司 | Greenhouse crop disease control method based on real-time feedback |
| CN112055079A (en)* | 2020-09-03 | 2020-12-08 | 华艺生态园林股份有限公司 | A cloud computing platform-based monitoring and early warning system for pests and diseases |
| CN112507869A (en)* | 2020-12-07 | 2021-03-16 | 广州博进信息技术有限公司 | Underwater target behavior observation and water body environment monitoring method based on machine vision |
| CN114342717A (en)* | 2021-12-27 | 2022-04-15 | 深圳市金阳盛城市服务集团有限公司 | Insect pest detecting and reporting method, device, equipment and storage medium |
| CN116258373A (en)* | 2023-03-15 | 2023-06-13 | 杭州盈禾嘉田科技有限公司 | Disease and pest detection, prediction and early warning system and method based on big data |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1462913A (en)* | 2003-05-15 | 2003-12-24 | 沈佐锐 | Intelligent system for monitoring healthy of ecology in greenhouse |
| CN100493212C (en)* | 2007-05-24 | 2009-05-27 | 中国科学院合肥物质科学研究院 | Intelligent early warning system of field crop diseases and insect pests |
| CN101938512A (en)* | 2010-08-13 | 2011-01-05 | 仲恺农业工程学院 | Crop disease and insect pest image information automatic monitoring system based on Internet of things |
| CN103139937A (en)* | 2013-03-12 | 2013-06-05 | 山东科技大学 | Crop disease and pest information real-time collection and interactive service technology and system based on mobile terminal |
| US20140259898A1 (en)* | 2013-03-15 | 2014-09-18 | Stephen Jens | Selectively eradicating plants |
| CN204482814U (en)* | 2015-03-17 | 2015-07-22 | 李玉杰 | A kind of autonomous check processing device of damage by disease and insect being applicable to green house of vegetables |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1462913A (en)* | 2003-05-15 | 2003-12-24 | 沈佐锐 | Intelligent system for monitoring healthy of ecology in greenhouse |
| CN100493212C (en)* | 2007-05-24 | 2009-05-27 | 中国科学院合肥物质科学研究院 | Intelligent early warning system of field crop diseases and insect pests |
| CN101938512A (en)* | 2010-08-13 | 2011-01-05 | 仲恺农业工程学院 | Crop disease and insect pest image information automatic monitoring system based on Internet of things |
| CN103139937A (en)* | 2013-03-12 | 2013-06-05 | 山东科技大学 | Crop disease and pest information real-time collection and interactive service technology and system based on mobile terminal |
| US20140259898A1 (en)* | 2013-03-15 | 2014-09-18 | Stephen Jens | Selectively eradicating plants |
| CN204482814U (en)* | 2015-03-17 | 2015-07-22 | 李玉杰 | A kind of autonomous check processing device of damage by disease and insect being applicable to green house of vegetables |
| Title |
|---|
| 田杰: "基于图像分析的小麦叶部病害识别方法研究", 《中国博士学位论文全文数据库信息科技辑》* |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110167344A (en)* | 2016-11-11 | 2019-08-23 | 生物生态服务提供商 | System and method for detecting flying animals |
| CN106780443A (en)* | 2016-11-30 | 2017-05-31 | 深圳前海弘稼科技有限公司 | The method for early warning of crop pest and the prior-warning device of crop pest |
| CN106779188A (en)* | 2016-11-30 | 2017-05-31 | 深圳前海弘稼科技有限公司 | Plant pest Forecasting Methodology and device in a kind of plantation equipment |
| CN106846462A (en)* | 2016-12-30 | 2017-06-13 | 北京农业信息技术研究中心 | Insect identifying device and method based on three-dimensional simulation |
| CN106846462B (en)* | 2016-12-30 | 2019-12-17 | 北京农业信息技术研究中心 | Insect recognition device and method based on three-dimensional simulation |
| CN107168367A (en)* | 2017-06-08 | 2017-09-15 | 中科院合肥技术创新工程院 | Accurate pesticide spraying system based on machine vision industry |
| CN107272620A (en)* | 2017-06-23 | 2017-10-20 | 深圳市盛路物联通讯技术有限公司 | A kind of method and device of the intelligent monitoring greenhouse based on Internet of Things |
| CN107301605A (en)* | 2017-06-29 | 2017-10-27 | 深圳前海弘稼科技有限公司 | The control method and device of crop pest |
| CN107357271A (en)* | 2017-06-30 | 2017-11-17 | 深圳前海弘稼科技有限公司 | The prevention and controls of chamber crop pest and disease damage, preventing and treating system |
| CN107357271B (en)* | 2017-06-30 | 2018-07-17 | 深圳春沐源控股有限公司 | The control method of chamber crop pest and disease damage, prevention system |
| WO2019126971A1 (en)* | 2017-12-25 | 2019-07-04 | 仲恺农业工程学院 | Pest disaster monitoring method based on machine vision |
| CN108040997A (en)* | 2017-12-25 | 2018-05-18 | 仲恺农业工程学院 | Insect pest monitoring method based on machine vision |
| US10729117B2 (en) | 2017-12-25 | 2020-08-04 | Zhongkai University Of Agriculture And Engineer | Pest monitoring method based on machine vision |
| CN108040997B (en)* | 2017-12-25 | 2020-09-11 | 仲恺农业工程学院 | A method of pest monitoring based on machine vision |
| IT201900002249A1 (en)* | 2019-02-15 | 2020-08-15 | Agrorobotica S R L | System and method of predicting the risk of insect attack |
| WO2020165701A1 (en)* | 2019-02-15 | 2020-08-20 | AGROROBOTICA S.r.l. | Insect attack risk prediction system and method |
| CN110427922A (en)* | 2019-09-03 | 2019-11-08 | 陈�峰 | One kind is based on machine vision and convolutional neural networks pest and disease damage identifying system and method |
| CN112034912A (en)* | 2020-09-03 | 2020-12-04 | 山东锋士信息技术有限公司 | Greenhouse crop disease control method based on real-time feedback |
| CN112055079A (en)* | 2020-09-03 | 2020-12-08 | 华艺生态园林股份有限公司 | A cloud computing platform-based monitoring and early warning system for pests and diseases |
| CN112055079B (en)* | 2020-09-03 | 2022-08-23 | 华艺生态园林股份有限公司 | Disease and pest monitoring and early warning system based on cloud computing platform |
| CN112507869A (en)* | 2020-12-07 | 2021-03-16 | 广州博进信息技术有限公司 | Underwater target behavior observation and water body environment monitoring method based on machine vision |
| CN112507869B (en)* | 2020-12-07 | 2024-04-09 | 广州博进信息技术有限公司 | Underwater target behavior observation and water environment monitoring method based on machine vision |
| CN114342717A (en)* | 2021-12-27 | 2022-04-15 | 深圳市金阳盛城市服务集团有限公司 | Insect pest detecting and reporting method, device, equipment and storage medium |
| CN116258373A (en)* | 2023-03-15 | 2023-06-13 | 杭州盈禾嘉田科技有限公司 | Disease and pest detection, prediction and early warning system and method based on big data |
| CN116258373B (en)* | 2023-03-15 | 2024-02-09 | 杭州盈禾嘉田科技有限公司 | Disease and pest detection, prediction and early warning system and method based on big data |
| Publication | Publication Date | Title |
|---|---|---|
| CN105850930A (en) | Early warning system and method of diseases and insect pests based on machine vision | |
| Nuske et al. | Automated visual yield estimation in vineyards | |
| RU2764872C2 (en) | Weed detection in natural environment | |
| Qing et al. | Automated counting of rice planthoppers in paddy fields based on image processing | |
| CN112837302B (en) | Method and device for monitoring state of die, industrial personal computer, storage medium and system | |
| CN102759528B (en) | Method for detecting diseases of crop leaves | |
| CN112347887B (en) | Object detection method, object detection device and electronic equipment | |
| CN102096808B (en) | Automatic Forecasting Method of Rice Planthopper Insect Situation | |
| CN101021939A (en) | Computer image processing-based crops disease diagnosing system | |
| CN103903252A (en) | Automatic cotton field pest situation monitoring device and automatic cotton field pest situation monitoring method | |
| CN105491271A (en) | Method and device for adjusting auxiliary light source in image collection | |
| WO2024051067A1 (en) | Infrared image processing method, apparatus, and device, and storage medium | |
| CN117058607A (en) | Plant growth state monitoring system based on image visual analysis | |
| CN109145848A (en) | A kind of wheat head method of counting | |
| CN114414742A (en) | Urban greening daily maintenance on-line intelligent monitoring management cloud system based on artificial intelligence | |
| KR20170114065A (en) | System and method for servicing controlment for crop image and growth information | |
| CN118396422A (en) | Intelligent peach blossom recognition and fruit yield prediction method based on fusion model | |
| CN105557175B (en) | Intelligent grass-removing and its control method | |
| CN116052141A (en) | Crop growth period identification method, device, equipment and medium | |
| CN107644200A (en) | Pest and disease damage gradient of infection detection method and system based on Android system | |
| CN103900498B (en) | A kind of cotton field automatic detection method of the growth of cereal crop seedlings and detection device thereof | |
| CN102640622B (en) | Cotton picker navigation information image detection method and system | |
| CN206270728U (en) | A Grape Disease Identification System | |
| CN118506162A (en) | Loquat growth period automatic identification method based on improved convolutional neural network | |
| CN114774253A (en) | Portable spore capturing and detecting device and detecting method thereof |
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | Application publication date:20160817 | |
| RJ01 | Rejection of invention patent application after publication |