KR20200044216A - System and method for predicting the occurrence of pests using Big Data - Google Patents

Abstract

Disclosed is a big data-based pest outbreak prediction system detecting and predicting signs of pest outbreak in a special crop. The system includes: a sensor unit collecting environment information on a special crop cultivation site; an imaging unit imaging special crop state information; a big data processing unit subject to cloud-based transmission of weather history information, special fruit cultivation information, a pest control manual, pest information, and pesticide information surrounding the special crop cultivation site; an information collection unit collecting information on the sensor unit, the imaging unit, and the big data processing unit; sensor information and imaging information databases respectively storing the sensor unit information and the imaging unit information collected at the information collection unit; a diagnosis execution unit executing a special crop pest diagnosis occurrence and pest control algorithm based on the information stored in the sensor information and imaging information databases and the information collected from the big data processing unit; and a user guide unit visualizing the result of the diagnosis execution unit by means of a geographic information system (GIS) and transmitting the result to a user interface. According to the present invention, a pest outbreak timing and a pest outbreak index depending on the correlation between environment information surrounding the cultivation site and the size of special crop fruit are predicted by means of poisson distribution and a neural network algorithm, and thus the accuracy of the prediction can be significantly improved and the occurrence (expansion) of damage to farmers can be significantly reduced.

Description

System and method for predicting the occurrence of pests using Big Data}

The present invention relates to a system and method for predicting pest outbreaks using big data that can minimize damage to pests and meteorological disasters through microclimate climate collection and image analysis through image information, forecast models, and the like.

In general, the Internet of Things (Internet of things) is a wired / wireless network device that is mounted on things in life or an environment in which objects in life are connected to a wired / wireless network to share information. Recently, items related to the Internet of Things have been widely developed. . By using the Internet of Things, it is possible to share information by connecting objects with networks in various fields such as home appliances and electronic devices, as well as healthcare, remote meter reading, smart homes, and smart cars.

On the other hand, in the field of agriculture, crop cultivation technology and management technology, etc. are steadily developing, but the damage of crops or fruit trees is still occurring due to various diseases, and the necessity of modernizing the cultivation technology of crops continues to increase.

Recently, IoT (Internet of Things) technologies that are rapidly developing are combined with various industries to develop new services. In the field of labor-intensive agricultural industry, IoT technology is converging to become more advanced. Using these advanced technologies, it is transforming to enable stable and sustainable growth in the agricultural industry that is dependent on nature.

In the nationwide production rate, pesticide outbreak prediction studies on special fruit trees (plums, citrons, etc.) that produce more than 30% in the Jeonnam region (Suncheon, Gwangyang, Hadong, Goheung, etc.) are underway, but the control manual being distributed is It is manufactured according to the average weather environment in Korea, so there is a limit to its practical application.

Although pest management of these crops is being implemented nationwide, independent systems for planting and controlling pests are often developed and operated, and in the event of a sudden pest, there is often a lack of a system for rapid data collection. In addition, there is no system to openly service the developed model by conducting a predictive model study for timely control of pests, so it is managed only as a result of the research project and is not standardized.

In addition, it identifies the occurrence of various diseases, pests, and physiological disorders in facility horticultural crops such as peppers, tomatoes, eggplants, and watermelons, collects multimedia information necessary for diagnosis in the field, and outlines important pests Density fluctuations, farming measures, symptom data, etc. are collected, and pests and physiological disorders that damage crops can be easily and accurately diagnosed at the farming unit, and tailored information suitable for the situation or farming environment that the farmers desperately need is needed. It is a difficult reality.

Republic of Korea Patent Publication No. 10-2018-0078698 (published on July 10, 2018)

The present invention has been devised to solve the above-mentioned conventional problems, and collects and applies real-time or periodic environmental information and micro-meteorological information in the micro-regional cultivation area, which is specialized in the wide-area weather information and agricultural environment index information, and the cultivation area. Therefore, it is an object of the present invention to provide a system and method for predicting the occurrence of pests using big data that can predict and spread the occurrence and spread of meteorological disasters and pests of special crops (plum, citron, etc.).

According to an aspect of the present invention, a system for detecting and predicting signs of pest occurrence of a special crop, comprising: a sensor unit for collecting environmental information of a special crop cultivation site; A photographing unit for photographing status information of the special crop; A big data processing unit that receives weather history information and special fruit cultivation information, pest control information, pest information, and pesticide information around the special crop cultivation area based on the cloud; An information collecting unit collecting information of the sensor unit, the photographing unit, and the big data processing unit; A sensor information and photographing information database in which sensor unit information and photographing unit information collected in the information collecting unit are respectively stored; A diagnosis performing unit for performing a pest diagnosis and control algorithm of a special crop based on information stored in the sensor information and photographing information database and information collected from the big data processing unit; And a user guide that visualizes the results of the diagnosis execution unit using a Geographic Information System (GIS) and transmits the result to the user interface.

The information collection unit, a streaming server that generates image information to deliver the information of the photographing unit in real time; A data acquisition server in which the sensor unit and the photographing unit information are temporarily stored; And a data web server that enables information stored in the data acquisition server to be checked on the web, and the diagnostic execution unit compares the information stored in the sensor information and the shooting information database with information collected from the big data processing unit. Diagnosis and control algorithms for pests of special crops can be performed.

Each of the sensor unit and the information collection unit may further include a wireless transmitter and a wireless receiver that wirelessly transmit and receive sensor information.

The diagnostic execution unit is equipped with a pest prediction software engine suitable for the growth environment of the special crop, a pest crop prediction software engine suitable for the cultivation environment, and is collected from a special crop plantation in a pest occurrence prediction model calculated based on the applied algorithm. Calculate a customized pest outbreak index by applying the measured data, but compare the values collected from the big data processing section with the measured values of the sensor section and the imaging section and calculate the deviation to generate possible pest information and pest outbreaks according to the range of deviation values Probability can be calculated.

The diagnostic execution unit may calculate a pest occurrence index according to a correlation between environmental information around a plantation and a fruit size of a special crop using a Poisson distribution, and apply a neural network to predict when a pest occurs.

According to another aspect of the present invention, as a method of detecting and predicting signs of pest occurrence of a special crop, (a) collecting environmental information of the special crop plantation, photographing growth information of the special crop, and weather history information around the special crop plantation And receiving special fruit cultivation information, control manual, pest information, pesticide information is transmitted in the cloud-based; (b) collecting environmental information and growth information of the special crop cultivation area using wireless transmission and reception; (c) comparing the environment information and growth information of the special crop cultivation field with information received based on the cloud to perform a pest diagnosis and control algorithm; And (d) transmitting the execution result of step (c) to a user interface, wherein step (c) is calculated based on an algorithm mounted on the pest prediction software engine suitable for the growth environment of the special crop. Calculate the customized pest outbreak index by applying the measured data collected from the special crop plantation to the outbreak forecast model, but compare the values collected from the big data processing section with the measured values of the sensor section and the imaging section and calculate the deviation to calculate the range of the deviation value Pest occurrence prediction method using big data is provided, which is characterized in that it is possible to generate pest information and pest probability.

According to the pest occurrence prediction system and method using the big data of the present invention described above, the user can easily monitor the growth environment of a special crop at any time, and receive information on the prediction of the occurrence of the pest, so that the possibility of the occurrence of the pest is quickly It is possible to minimize damage and spread by proceeding with follow-up.

In addition, by using the Poisson distribution and neural network algorithm to predict the pest incidence index and the timing of the pest incidence according to the correlation between the environment information of the cultivation area and the size of the special crop, the accuracy is significantly improved, resulting in damage to farmers ( Diffusion) can be greatly reduced.

1 is a diagram showing a pest occurrence prediction system using big data according to an embodiment of the present invention,
Figure 2 is a block diagram showing a pest generation prediction system using big data according to an embodiment of the present invention,
3 is a diagram showing a GUI in a pest occurrence prediction system using big data according to an embodiment of the present invention,
4 is a flowchart illustrating a method for predicting pest occurrence using big data according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in more detail with reference to the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various different forms, and only the present embodiments allow the disclosure of the present invention to be complete, and the scope of the invention to those skilled in the art is completely It is provided to inform you. The same reference numerals in the drawings refer to the same elements.

The pest occurrence prediction system and method using big data according to a preferred embodiment of the present invention provides prediction information on pests and meteorological disasters in real time through cloud-based big data processing, and can be utilized in various environments through an open service interface In order to enable, the weather conditions around the planting area of special crops and the control manual, pesticide information, pest information, and crop planting information of the applied crops are delivered in the cloud-based and compared with the underground / ground environmental information around the plantation, pests, Provides forecast information for meteorological disasters. In other words, the present invention collects and applies real time / ground environmental information and fine weather information in a wide area weather information, agricultural environment index information, and micro area cultivation areas specialized in the cultivation area in real time, and causes meteorological disasters of special crops (plum, citron, etc.). , It is prepared to prevent the spread of damage early and increase the harvest by predicting the signs and spread of disease and pests and delivering them to users in real time.

Hereinafter, the present invention will be described in detail with reference to Examples.

As illustrated in FIG. 1, the pest occurrence prediction system using big data according to an embodiment of the present invention is a system for detecting and predicting signs of pest occurrence of a special crop, the sensor unit 100, the photographing unit 200, It includes a big data processing unit 250, an information collection unit 300, sensor information and photographing information databases 400,410, a diagnosis performing unit 500, and a user guide unit 600.

First, the sensor unit 100 collects environmental information of a special crop cultivation site, and is installed in a greenhouse where the special crop is grown to detect growth information. Specifically, the sensor unit 100 may measure temperature, humidity, and illuminance inside a plastic greenhouse where plums and citrons grow, and may measure EC (Electrical Conductivity) of the soil.

The photographing unit 200 is a kind of CCTV that photographs the status information of a special crop, and is similarly installed inside the vinyl house.

The big data processing unit 250 receives weather history information and special fruit cultivation information, a control manual, pest information, and pesticide information around the special crop cultivation area based on the cloud.

Specifically, the big data processing unit 250 is based on the data of related organizations such as the Korea Meteorological Administration, the past and present weather history information around the special crop plantation, agricultural environment index information through XML transmission, FTP transmission, etc. in real time or periodic Is sent to. In addition, the big data processing unit 250 receives real-time or periodic transmission of special crop cultivation information, pest information that may occur on the special crop, pest control manual, pesticide control information for pest control, etc., through agricultural organizations and the Internet. To be more specific, the big data processing unit 250 is, for example, various temperature, humidity, illuminance, EC (electrical conductivity) in a range of conditions, such as plums, citrons, pest occurrence probability, possible pests, optimal growth environment information, etc. Information can be sent.

Next, the information collection unit 300 collects information of the sensor unit 100, the photographing unit 200, and the big data processing unit 250.

As shown in FIG. 1, the information collection unit 300 includes a streaming server 310, a data acquisition server 320, and a data web server 330.

In addition, the present invention further includes a wireless transmitting unit 350 and a wireless receiving unit 360 that are respectively provided in the sensor unit 100 and the information collecting unit 300 to wirelessly transmit and receive the sensing information of the sensor unit 100. Specifically, the wireless transmission unit 350 is provided in the vinyl house and receives the detection signal of the sensor unit 100 and periodically transmits it to the wireless reception unit 360, and the wireless reception unit 360 is outside the vinyl house (around the area where the information collection unit is installed) Etc.) to receive transmission information of the wireless transmitter and the received information is stored in the data acquisition server 320. For example, the wireless transmitter 350 may transmit in a period of 1 hour to 6 hours, and wireless transmission and reception is possible using a wireless communication network such as the Internet, WIFI, LTE.

First, the streaming server 310 generates video information to deliver information of the photographing unit 200 as a real-time video, and is provided to provide data transmitted from the photographing unit 200 as a video streaming service, and a user uses a smartphone. By using a wireless communication device, such as can be checked in real time the captured image of the imaging unit 200.

The data acquisition server 320 is temporarily stored as the information of the sensor unit 100 and the photographing unit 200, and the stored information is finally transmitted to and stored in the sensor information and the photographing information databases 400 and 410 to be described later.

The data web server 330 is provided so that information stored in the data acquisition server 320 can be checked on the web, and the user guide unit 600 described later uses the information stored in the data web server 330 in a user interface (smartphone, etc.). ), So that the user can check the sensing value of the sensor unit as well as the captured image of the shooting unit in real time.

Next, the sensor information and the photographing information databases 400 and 410 are separately stored in the information collecting unit 300, specifically, the sensor unit information and the photographing unit information primarily collected / stored in the data acquisition server 320, respectively.

Next, the diagnosis execution unit 500 is based on the information stored in the sensor information and the shooting information database 400 and 410 and the information collected in real time or periodically through the big data processing unit 250 to diagnose and control the pest of the special crop. To perform. The diagnosis execution unit 500 compares the information stored in the sensor information and the shooting information databases 400 and 410 with the information collected through the big data processing unit 250 to perform a pest diagnosis and control algorithm of a special crop.

Specifically, the diagnosis execution unit 500 is equipped with a pest prediction software engine suitable for the growing environment of the special crop. In particular, the pest crop prediction software engine suitable for the weather environment in the cultivation area (around the vinyl house) is mounted and calculated based on the applied algorithm. The customized pest outbreak index is calculated by applying the measured data (temperature, humidity, illuminance, EC, etc.) collected from the special crop cultivation model to the predicted pest outbreak model.

The pest prediction software engine is a cloud-based database of various information such as weather history and special fruit cultivation information, control manuals, pest information, pesticide information, sensor information, and pesticide information collected from the big data processing unit 250 collected from the big data processing unit 250 By comparing the values stored in (400,410) and calculating the deviation, the possible pest information and the probability of occurrence of the pest according to the range of the deviation value are calculated. Here, the pest prediction software engine calculates the pest occurrence index separately according to crop types such as plums and citrons. To this end, the big data processing unit 250 receives pest-related information, optimal growth environment information, etc. for various special crops. The prediction software engine calculates the pest outbreak index based on the information corresponding to the special crop.

In an embodiment of the present invention, the diagnosis performing unit 500 calculates the pest occurrence index according to the correlation between the environment information around the plantation and the fruit size of the special crop using the Poisson distribution algorithm, and applies the neural network To predict when pests will occur.

Here, the Poisson distribution algorithm is an algorithm using the Poisson distribution, which is a mathematical statistical technique, and uses a discrete probability distribution representing how many times an event will occur within a unit time. When the expected value of the number of times an event will occur within a predetermined time is called a lambda, the probability that the event will occur n times is as follows.

[Mathematics]

Where e is the natural constant

Artificial neural network (ANN) is a statistical learning algorithm inspired by the neural network of biology (especially the brain of the animal's central nervous system) in machine learning and cognitive science. The artificial neural network refers to a model in which artificial neurons (nodes) that form a network through synaptic binding have a problem-solving ability by changing the strength of synaptic binding through learning. The diagnosis execution unit 500 calculates the probability of occurrence of a pest using the Poisson distribution algorithm and the neural network algorithm described above, converts it into a pest occurrence index, and reflects information of the big data processing unit 250 in various ways to generate a pest Predict.

The present invention significantly improves the accuracy by predicting the pest incidence index and the timing of the pest incidence according to the correlation between the environment information around the plantation and the fruit size of the special crop using the Poisson distribution and the neural network algorithm. The occurrence (diffusion) can be greatly reduced.

Next, the user guide unit 600 transmits the results of the diagnosis performing unit 500 to the user interface so that the user generates real-time pests measured in the diagnosis performing unit 500 as well as the actual data and photographed images of the special crop plantation in real time. The index and the time of occurrence of the pest are received to help predict and prepare for the signs of the occurrence of the pest in advance. The user guide 600 may transmit to the GUI of the user's smartphone through an Internet network, Wi-Fi, Bluetooth, or the like.

In an embodiment of the present invention, as shown in FIG. 3, the user guide unit 600 visualizes the result of the diagnosis performing unit 500 using a Geographic Information System (GIS) and transmits it to the user interface.

On the user interface screen, such as a smartphone, the level of the pest notification stage is color-coded based on the map, and possible or occurring pests (e.g., plum black star disease) along with the warning notification stage, warning delivery time, and warning notification contents ), When the occurrence of the pest, pest description, control methods, etc. are provided based on the map image, so that the user can quickly and accurately follow up.

As shown in FIG. 4, the method for predicting the occurrence of pests using big data according to an embodiment of the present invention is a method for detecting and predicting signs of occurrence of pests of a special crop, (a) collecting environmental information of a special crop plantation And photographing growth information of the special crop and receiving weather history information and special fruit cultivation information, a control manual, pest information, and pesticide information around the special crop plantation in a cloud-based manner; (b) collecting environmental information and growth information of the special crop cultivation area using wireless transmission and reception; (c) comparing the environment information and growth information of the special crop cultivation field with information received based on the cloud to perform a pest diagnosis and control algorithm; And (d) transmitting the result of the step (c) to the user interface.

First, through the sensor unit 100 and the photographing unit 200, environmental information (temperature, humidity, illuminance, EC, etc.) of the special crop cultivation site and the growth of the special crop are photographed, and the special crop is processed through the big data processing unit 250. The weather history information around the cultivation area, special fruit cultivation information, control manual, pest information, and pesticide information are received in the cloud (S100).

Next, the cultivation environment information and photographing information collected in step S100 are transmitted to the information collection unit 300 using the wireless transmission unit 350 and the wireless reception unit 360 (S200).

The information collected in the next step S200 and the special data cultivation information, control manual, pest information, pesticide information, etc. collected in real time or periodically through the big data processing unit 250 are compared and analyzed to perform pest diagnosis and control algorithm ( S300).

In step S300, actual data (temperature, humidity, illuminance, EC, etc.) collected from the special crop cultivation in the pest outbreak prediction model calculated based on the algorithm applied using the pest prediction software engine suitable for the growth environment of the special crop (especially the cultivation environment) ) And a variety of information collected through the big data processing unit 250 to calculate a customized pest outbreak index. Specifically, in step S300, using the Poisson distribution algorithm, the pest incidence index according to the correlation between the environment information around the plantation and the fruit size of the special crop is calculated, and the information collected by the big data processing unit, the sensor unit and the imaging Predict the time when pests occur by applying various information and neural networks.

By sending the pest information, control method, etc. for the pest that has occurred or has a high probability of occurrence, along with the pest outbreak index and the probability of occurrence of the pest calculated in the next step S300, the user checks the prediction information related to the outbreak in real time. To prepare for this (S400).

Although the present invention has been described with reference to the accompanying drawings and the preferred embodiments described above, the present invention is not limited thereto, and is limited by the claims below. Therefore, those of ordinary skill in the art can variously modify and modify the present invention without departing from the technical spirit of the claims to be described later.

100: sensor unit 200: shooting unit
250: Big data processing unit 300: Information collection unit
310: streaming server 320: data acquisition server
330: data web server 400: sensor information database
410: photographing information database 500: diagnosis performing unit
600: user guide

Claims (6)

As a system for detecting and predicting signs of pest occurrence of special crops,
A sensor unit 100 that collects environmental information of a special crop plantation;
A photographing unit 200 for photographing status information of a special crop;
A big data processing unit 250 that receives weather history information and special fruit cultivation information, pest control information, pest information, and pesticide information around the special crop cultivation area based on the cloud;
An information collecting unit 300 for collecting information of the sensor unit, the photographing unit, and the big data processing unit;
A sensor information and photographing information database (400,410) in which sensor unit information and photographing unit information collected in the information collecting unit are respectively stored;
A diagnosis performing unit 500 for performing a diagnosis and control algorithm for pests and pests of special crops based on information stored in the sensor information and shooting information database and information collected from the big data processing unit; And
Pest occurrence prediction system using big data including a user guide unit (600) that visualizes the results of the diagnosis execution unit using a GIS (Geographic Information System) and transmits it to a user interface. According to claim 1,
The information collecting unit 300,
A streaming server 310 that generates image information to deliver information of the photographing unit 200 as a real-time image;
A data acquisition server 320 in which the sensor unit 100 and the photographing unit 200 information are temporarily stored; And
And a data web server 330 that enables information stored in the data acquisition server 320 to be checked on the web.
The diagnostic execution unit 500 compares the information stored in the sensor information and the photographing information database with information collected from the big data processing unit, and performs big data diagnosis and control algorithms of pests of special crops. Pest outbreak prediction system using. According to claim 1,
A pest generation prediction system using big data, characterized in that it further includes a wireless transmitter (350) and a wireless receiver (360) provided for each of the sensor unit and the information collecting unit to wirelessly transmit and receive sensor information. According to claim 1,
The diagnosis performing unit 500,
Mounting the pest prediction software engine suitable for the growth environment of the special crop and mounting the pest crop prediction software engine suitable for the meteorological environment and applying the measured data collected from the special crop plantation to the pest occurrence prediction model calculated based on the applied algorithm. To calculate the customized pest outbreak index, but compares the values collected from the big data processing unit 250 with the measured values of the sensor section and the imaging section and calculates the deviation to determine possible pest information and probability of occurrence of the pest according to the range of the deviation value Pest occurrence prediction system using big data characterized by calculating. According to claim 4,
The diagnosis performing unit 500,
Using the Poisson distribution, we calculate the pest incidence index according to the correlation between the environment information around the cultivation area and the size of the special crop, and apply the neural network to predict when the pest will occur. Pest outbreak prediction system. As a method of detecting and predicting signs of pest occurrence of special crops,
(a) collecting environmental information of the special crop cultivation site, photographing growth information of the special crop cultivation, and receiving weather history information and special fruit cultivation information, a control manual, pest information, and pesticide information around the special crop cultivation in a cloud-based manner;
(b) collecting environmental information and growth information of the special crop cultivation area using wireless transmission and reception;
(c) performing the pest diagnosis and control algorithm by comparing the environment information and growth information of the special crop with the information received based on the cloud; And
(d) transmitting the result of step (c) to a user interface,
In the step (c), a customized pest outbreak index is calculated by applying actual data collected from a special crop cultivation model to a pest outbreak prediction model calculated based on an algorithm mounted in the pest prediction software engine suitable for the growth environment of the special crop, Big data, characterized in that it compares the values collected from the big data processing unit 250 with the measured values of the sensor unit and the imaging unit, calculates the deviation, and calculates possible pest information and probability of occurrence of the pest according to the range of the deviation value. Prediction method using pests.

Images