KR102187654B1 - Low altitude drone and Crop Cultivating Information Acquisition System including the same - Google Patents

Abstract

The present invention provides a crop cultivation information obtaining system, which includes: a control center generating a flight path in accordance with a lot of land formed in consideration of predetermined land information and take-off and landing points; and a low altitude unmanned aerial vehicle obtaining environmental information or low altitude crop images while performing low altitude flight along the flight path, wherein the control center analyses the low altitude crop images to obtain crop information and displays the crop information in accordance with a coordinate to generate a crop map.

Description

Low altitude drone and Crop Cultivating Information Acquisition System including the same}

The present invention relates to a method and system for surveying the cultivation area of crops and trees, and in particular, to a method and system for acquiring information such as crop cultivation area and crop type using a drone.

The content described in this section merely provides background information on the present embodiment and does not constitute the prior art.

As the variability in the supply and demand of crops increases due to extreme weather, change of cropland use, crop renewal, changes in farming methods, and scale-up of cultivation due to global warming, precise agricultural observations are required. Statistics that grasp the current status of crops and trees that actually exist in agricultural land and forestry land are basic information for predicting the supply and demand of agricultural products and forest products and establishing future supply and demand plans. However, the existing methods are limited to the method of constructing spatial information based on building technology or surveying technology, and thus, there is insufficient limit as a data for grasping vegetation at the actual site. In other words, since it is not possible to provide accurate information on the current vegetation grasping when the existing method is followed, there is a limitation that the future demand forecast is also inaccurate.

In the past, institutions such as Rural Economic Research Institute and Statistics Korea provided information on vegetation in farmland and mountain areas, but there is a reason that the forecast is inaccurate for the above-described reasons, so there is too much supply of specific crops and various problems such as a plunge in prices. This is because the basic area of land and cultivation area itself, which is the basis of demand forecasting, was wrong.

The embodiments of the present invention have been devised to solve the above problems, and use a low-altitude drone to increase observation accuracy, and to reinforce the supply and demand prediction power by quickly and quickly grasping the cultivation area shipment fluctuations and extreme weather conditions. There is a purpose.

Still other objects, not specified, of the present invention may be additionally considered within the range that can be easily deduced from the following detailed description and effects thereof.

According to an aspect of the present embodiment, the present invention provides a control center that generates a flight path according to a lot formed in consideration of preset terrain information, take-off point and landing point, and performs low altitude flight according to the flight path, and provides environmental information or low altitude. Crop cultivation information, comprising a low-altitude unmanned aerial vehicle that acquires a crop image, wherein the control center analyzes the low-altitude crop image to obtain crop information, and displays the crop information according to coordinates to generate a crop map. Propose an acquisition system.

According to another embodiment of the present invention, the present invention provides a driving unit that generates a driving force that enables flight along a flight path generated in consideration of preset terrain information, a sensor unit that acquires surrounding environment information, and the flight path. An image acquisition unit that performs a low altitude flight and acquires a low altitude crop image according to, a processor that obtains crop information by analyzing the low altitude crop image, and displays the crop information according to coordinates to generate a crop map, and the obtained low altitude A low-altitude unmanned aerial vehicle including a communication unit for transmitting an image or an operation result in the processor to an external communication device is proposed.

As described above, according to the embodiments of the present invention, the present invention enables a low-altitude unmanned aerial vehicle performing a low-altitude flight to obtain a low-altitude crop image obtained at a low altitude, and using the obtained low-altitude crop image , There is an effect that image processing is possible for the characteristics of crops.

In addition. According to the embodiments of the present invention, the present invention enables object recognition of the low-altitude crop image obtained in this way through an artificial neural network, so that the current status of crops for each lot can be more accurately identified compared to the existing technology.

Even if it is an effect not explicitly mentioned herein, the effect described in the following specification expected by the technical features of the present invention and the provisional effect thereof are treated as described in the specification of the present invention.

1 is a block diagram schematically showing a system for obtaining crop cultivation information according to an embodiment of the present invention.
2 and 3 are exemplary views showing a low-altitude unmanned aerial vehicle according to an embodiment of the present invention.
4 to 9 are exemplary diagrams illustrating a process of obtaining crop cultivation information according to an embodiment of the present invention.
10 is a view showing a state of a crop by using an image acquired through a low-altitude drone flight according to an embodiment of the present invention.
11 is a diagram illustrating extraction of an orthogonal image through low-altitude terrain flight according to an embodiment of the present invention.
12 to 14 are flowcharts illustrating a process performed by the crop cultivation information acquisition system to acquire crop cultivation information according to an embodiment of the present invention.
15 is a block diagram illustrating a learning concept for crop analysis performed in an AI server according to an embodiment of the present invention.
16 is a conceptual diagram illustrating a method of generating an input image for image classification through a generator determined through FIG. 15.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. Advantages and features of the present invention, and a method of achieving them will become apparent with reference to the embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments to be posted below, but may be implemented in a variety of different forms, and only these embodiments make the posting of the present invention complete, and common knowledge in the technical field to which the present invention pertains. It is provided to completely inform the scope of the invention to those who have it, and the invention is only defined by the scope of the claims. The same reference numerals refer to the same components throughout the specification.

Unless otherwise defined, all terms (including technical and scientific terms) used in the present specification may be used as meanings that can be commonly understood by those of ordinary skill in the art to which the present invention belongs. In addition, terms defined in a commonly used dictionary are not interpreted ideally or excessively unless explicitly defined specifically.

The terms used in the present application are only used to describe specific embodiments, and are not intended to limit the present invention. The singular expression includes a plurality of expressions unless the context clearly indicates otherwise. In the present application, terms such as "comprise" or "have" are intended to designate the presence of features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, but one or more other features. It is to be understood that the presence or addition of elements or numbers, steps, actions, components, parts, or combinations thereof, does not preclude in advance.

Terms including ordinal numbers, such as second and first, may be used to describe various elements, but the elements are not limited by the terms. These terms are used only for the purpose of distinguishing one component from another component. For example, without departing from the scope of the present invention, a second component may be referred to as a first component, and similarly, a first component may be referred to as a second component. The term and/or includes a combination of a plurality of related listed items or any of a plurality of related listed items.

The present invention relates to a low-altitude unmanned aerial vehicle and a crop cultivation information acquisition system including the same.

The crop cultivation information acquisition system 10 can quickly and accurately determine the cultivation area using the low-altitude unmanned aerial vehicle 100 and the control center 200. The crop cultivation information acquisition system 10 enables fast and accurate observations, can bring economic effects, and can increase the quantity and quality of observations. Accordingly, farmers' income can be guaranteed, consumer prices can be stabilized, agricultural observation accuracy can be improved, and supply and demand forecasting power can be strengthened.

Here, rapid observation can derive rapid results for conversion of crops or supply and demand, and accurate observation enables the advancement and modernization of inaccurate observation data, and economic observation reduces the cost of observation and economic damage caused by observation errors. There is an effect. In addition, the quantity and quality of observations has the effect of providing various information such as meteorological disasters, pests, and vegetation index.

The crop cultivation information acquisition system 10 may investigate high altitude crops, and acquire images through a drone to quickly and accurately observe crop cultivation area and growth. The crop cultivation information acquisition system 10 can be observed in an area of 850 m or more above sea level, where the topography is rugged and the daily temperature difference is severe.

According to an embodiment of the present invention, the crop cultivation information acquisition system 10 may grasp and display a change pattern of highland cabbage cultivation area by time and sequentially through drone photography. Through fast and accurate processing of images acquired using drones and time series change management based on a geographic information system (GIS), the efficiency of observation work can be improved, and drone image data can be effectively utilized in overall agricultural observation work. Here, the highland cabbage representing the crop is an example and is not necessarily limited thereto.

Geographic Information System (GIS) is a type of information system that processes and manages geographical information, that is, spatially distributed information. Here, spatially distributed information, that is, geographic information, is usually displayed in the form of a map, and the geographic information system refers to an information system that can input, store, process, analyze, manage, and output such geographic information. GIS can derive parcel division and area based on Farm Map. Here, the Farm Map may be used when extracting a boundary line for each lot.

Therefore, the crop cultivation information acquisition system 10 can observe the time-series change of the cultivation area, can check the economical efficiency and effectiveness of observing the cultivation area and shipment fluctuation through a drone, and can identify variability through real-time image identification. , It is possible to preemptively secure basic data for growth judgment, such as effectively observing open field crops, and grasping pests and weather damage conditions for advancement.

1 is a block diagram schematically showing a system for obtaining crop cultivation information according to an embodiment of the present invention.

Referring to FIG. 1, the crop cultivation information acquisition system 10 includes a low-altitude unmanned aerial vehicle 100 and a control center 200. The crop cultivation information acquisition system 10 may omit some of the various components exemplarily illustrated in FIG. 1 or may additionally include other components.

According to an embodiment of the present invention, the crop cultivation information acquisition system 10 may generate information by performing a flight through the low-altitude unmanned aerial vehicle 100. The low-altitude unmanned aerial vehicle 100 may acquire an image, a thermal image, a hyperspectral image, and the like and transmit it to the control center 200. The control center 200 generates and demonstrates the flight path of the low-altitude unmanned aerial vehicle 100, collects information produced by the low-altitude unmanned aerial vehicle 100, and processes big data to classify AI crops, cultivate, and derive the current state of harvest. Visualization, GIS-based analysis and expression, etc. can be performed, and the flight path, collection method development (DSM 3D terrain input, Farm Map input), survey area, transmission data transmission and reception, etc. can be managed by using spatial information. In addition, the crop cultivation information acquisition system 10 can provide the route provision, collection, storage, and analysis of the control center 200 to the low-altitude unmanned aerial vehicle 100, public institutions, research institutions, companies, academia, and other industries. However, it is not necessarily limited thereto.

The low-altitude unmanned aerial vehicle 100 performs low-altitude flight according to the flight path and acquires environmental information or low-altitude crop images. Here, the flight path is a path through which the low-altitude unmanned aerial vehicle 100 performs low-altitude flight and obtains environmental information or a low-altitude crop image, and a path for flying in consideration of a take-off point and a landing point may be generated.

The low-altitude unmanned aerial vehicle 100 includes a sensor unit 110, a driving unit 120, an image acquisition unit 130, a processor 140, and a communication unit 150. The low-altitude unmanned aerial vehicle 100 may omit some components from among the various components exemplarily illustrated in FIG. 1 or may additionally include other components.

The sensor unit 110 may acquire the surrounding environment information.

The sensor unit 100 includes a front sensor 112 that detects an obstacle located in front, a lower sensor 114 that detects an obstacle located below, an ultrasonic sensor 116 that detects surrounding obstacles through ultrasonic waves, and It may include a proximity sensor 118 for detecting an obstacle, but is not limited thereto. Here, obstacles represent blocking and delaying the movement of the drone.

The sensor unit 100 may acquire environmental information including at least one of obstacle information detected by the front sensor 112, the lower sensor 114, the ultrasonic sensor 116, and the proximity sensor 118.

The driving unit 120 may generate a driving force that enables the flight of the low-altitude unmanned aerial vehicle 100.

The image acquisition unit 130 may move along the flight path and acquire a low-altitude crop image.

The processor 140 may reset a flight path in consideration of environmental information and the acquired low-altitude crop image, and control the driving unit and the image acquisition unit.

When an obstacle detected according to the acquired environmental information is located in the flight path, the processor 140 re-analyzes the movement path for moving to the next point in consideration of the number and size of the detected obstacles, and resets the flight path to reset the path. And may transmit the reset path to the control center 200.

Specifically, the reset path is generated by acquiring surrounding environment information while the low-altitude unmanned aerial vehicle 100 moves along the flight path generated by the control center 200, and resetting the optimal flight path based on the acquired environment information. I can. Here, the environmental information may include elements that interfere with the flight of the low-altitude unmanned aerial vehicle 100, such as the number of obstacles, the size of the obstacle, and weather.

The processor 140 may perform primary crop recognition by analyzing the first low-altitude crop image acquired by the image acquisition unit 130 set to face directly downward. When the crop information cannot be obtained through the primary crop recognition, the processor 140 receives the second low altitude crop image obtained by adjusting the angle of the image acquisition unit 130 set to face directly downward to a preset angle. By analyzing, secondary crop recognition can be performed. At this time, the processor 140 generates a composite image by synthesizing the first low-altitude crop image and the second low-altitude crop image through a hostile generation neural network (GAN) when it is not possible to obtain crop information through secondary crop recognition. By analyzing the synthesized image, the third crop recognition is performed.

According to an embodiment of the present invention, the processor 140 transmits the first low-altitude crop image, the second low-altitude crop image, and the composite image to the control center 200 when crop information cannot be obtained through the third crop recognition. And, the low-altitude unmanned aerial vehicle can move to the next point.

Here, the control center 200 performs crop recognition by analyzing the first low-altitude crop image, the second low-altitude crop image, and the composite image, and transmits a re-taking signal to the low-altitude unmanned aerial vehicle 100 when crop recognition is not performed. I can.

The communication unit 150 may transmit the obtained environmental information, low-altitude crop image or crop information to the control server 200 and receive a flight path.

The control center 200 generates a flight path according to a lot formed in consideration of a take-off point and a landing point according to preset terrain information.

The control center 200 analyzes low-altitude crop images to obtain crop information, and displays crop information according to coordinates to generate a crop map.

The control center 200 includes a route setting unit 210 and a crop detection unit 220. The crop detection unit 220 includes an image analysis unit 222, an area calculation unit 224, a coordinate extraction unit 226, and a map generation unit 228. The control center 200 may omit some of the various components exemplarily illustrated in FIG. 1 or may additionally include other components.

The route setting unit 210 extracts a representative point of the parcel based on preset topographic information including parcel information about the parcel, and may set the flight path of the low-altitude unmanned aerial vehicle 100 in consideration of the representative point of the parcel. .

The path setting unit 210 may generate a flight path in consideration of the flight trajectory according to the extracted representative point indicating the center point of the parcel. The lot information includes divisional boundary information between lots and elevation information of areas included in the lot.

The crop detection unit 220 may generate a crop map by analyzing the low-altitude crop image acquired according to the flight path.

The image analysis unit 222 may generate crop information by analyzing low-altitude crop images and classifying crops.

The area calculation unit 224 may calculate a cultivation area according to crop information among the total area according to the flight path.

The coordinate extraction unit 226 may extract coordinates of crop information.

The map generator 228 may generate a crop map using flight paths, crop information, cultivation area, and coordinates.

2 and 3 are exemplary views showing a low-altitude unmanned aerial vehicle according to an embodiment of the present invention.

FIG. 2 is a view showing the structure of a low-altitude unmanned aerial vehicle according to an embodiment of the present invention, and FIG. 3 is a view showing a structure of a rear-facing structure of a low-altitude unmanned aerial vehicle according to an embodiment of the present invention.

The low-altitude unmanned aerial vehicle 100 of the present embodiment may be implemented as a drone, a small helicopter, or an airplane, and specifically, it may be implemented as a vehicle that moves mainly by remote control without a person boarding and controlling it.

The low-altitude unmanned aerial vehicle 100 may be implemented as a housing body and a flight unit.

The housing body forms the body skeleton of the low-altitude unmanned aerial vehicle 100.

The flight unit may be implemented with an arm portion and a wing portion, and may be formed of four or six pairs, as shown in FIGS. 2 and 3. The flight unit may be included in the driving unit 120 of the low-altitude unmanned aerial vehicle 100.

The driving unit 120 may drive various actuators included in the low-altitude unmanned aerial vehicle 100. Here, the actuator is a device that performs the operation required for the low-altitude unmanned aerial vehicle 100 through motor drive, gear drive, etc., for example, a drive motor for driving the propellers 122a, 122b, 122c, 122d of the flight unit have. The driving unit 120 may generate a control signal for driving it.

Accordingly, the driving unit 120 may be driven to move along the flight path transmitted from the control center 200.

The processor 140 may recognize crops through processing of the image signal acquired by the low-altitude unmanned aerial vehicle 100.

The communication unit 150 transmits an image obtained through communication with an external server to the outside or transmits the result information determined by the processor 140 to the outside, and receives a control signal from an external operation unit. I can.

The low-altitude unmanned aerial vehicle 100 may further include a memory (not shown) and a storage unit (not shown). The memory may temporarily store the signal processed by the processor 140. The storage unit is a nonvolatile auxiliary storage device that stores information stored in the memory.

Referring to FIG. 3, the low-altitude unmanned aerial vehicle 100 includes a sensor unit 110 and an image acquisition unit 130 on a rear surface thereof.

The image acquisition unit 130 includes a camera 132 and a gimbal 134.

The camera 132 acquires an image of a crop through low altitude flight.

The gimbal 134 connects the low-altitude unmanned aerial vehicle 100 and the camera 132 and can rotate about a direction axis. The gimbal 134 grips the camera and controls the gaze direction of the camera. Here, the processor 140 may control the driving of the gimbal 134 and consequently, the gaze direction of the camera 132.

Mainly, the camera 132 faces directly downward, but when crops cannot be distinguished by only the ground image, it is necessary to acquire an image in a direction of 40 to 50 degrees lateral to a crop such as corn.

It is preferable that the gimbal 134 has a 3-axis gimbal structure, and it is possible to shoot an image in a target direction without shaking when flying in turbulence and strong winds.

The target angle of the gimbal 134 may be set by the processor 140 of the low-altitude unmanned aerial vehicle 100. Here, the target angle may be set to look directly underneath, and may be implemented at 180 degrees based on the low-altitude unmanned aerial vehicle 100.

When the target angle is set, the processor 140 may continuously correct the angle so that the set angle is not deformed even if the low-altitude unmanned aerial vehicle 100 receives an external impact. For example, the target angle is set to 180 degrees as an angle looking directly below, but when the low-altitude unmanned aerial vehicle 100 is skewed to the left or right, it may be smaller or larger than 180 degrees, and the camera 132 is directly below It can be corrected by adding an angle that is skewed to the left or right at 180 degrees to look at.

Specifically, the processor 140 primarily performs crop recognition by processing the low-altitude crop image acquired by the image acquisition unit 130, and when the crop is not recognized, the image acquisition unit 130 looking directly below. ) Is adjusted to a preset angle to acquire low-altitude crop images, and then crop recognition is performed secondarily. At this time, the processor 140 performs crop recognition by sequentially adjusting the angle of the image acquisition unit 130, and if the crop is not recognized, sends a low-altitude crop image to the control center 200 for reconfirmation. At this time, when the crop is not read according to the low-altitude crop image even in the control center 200, a re-photographing signal for additional photographing may be transmitted to the low-altitude unmanned aerial vehicle 100 to retake the parcel.

The low-altitude unmanned aerial vehicle 100 may capture a low-altitude crop image from a current position or a low-altitude crop image by moving a predetermined distance when the camera 132 moves by a preset angle in a direction facing directly below.

According to an embodiment of the present invention, when the camera 132 moves by a preset angle in a direction looking directly below, the low altitude crop image after moving by a distance proportional to the preset angle. You can shoot. At this time, the moving path may move to a path moving from the current position to the next path in consideration of the flight path in which the low-altitude unmanned aerial vehicle 100 is set.

The sensor unit 110 includes a front sensor 112, a lower sensor 114, an ultrasonic sensor 116, and a proximity sensor 350.

The front sensor 112 is preferably a stereo image sensor. The front sensor 112 acquires a stereo image for detecting an obstacle located in front through stereo image processing. An obstacle in front may be detected through image processing in a processor described below or an image processing processor for image processing.

The lower sensor 114 is preferably a stereo image sensor. The downward sensor 114 is a sensor for recognizing an obstacle located in a downward direction through stereo image processing.

The ultrasonic sensor 116 detects surrounding obstacles through ultrasonic signals. The image sensor is not suitable for night flight, but the ultrasonic sensor 116 can detect obstacles even at night.

The proximity sensor 350 may detect an obstacle in proximity. For example, the proximity sensor 350 is preferably an infrared proximity sensor, but is not limited thereto.

Data acquired through the sensor unit 110 and the image acquisition unit 130 may be converted into signals that can be processed by the processor 140 when transmitted to the processor 140.

4 to 10 are exemplary diagrams illustrating a process of obtaining crop cultivation information according to an embodiment of the present invention.

4 is a diagram sequentially illustrating a process of obtaining crop cultivation information according to an embodiment of the present invention.

4A is a process of generating a flight path in the control center 200. At this time, the generated flight path is transmitted to the unmanned aerial vehicle 100. For example, in an area where a wind power plant is located, a flight path can be set by designing to deviate from the path so that it is outside the sphere of influence of the magnetic field. In this way, the flight path is set in consideration of all factors that may interfere with the flight of the unmanned aerial vehicle 100. The flight route extracts the plane coordinate of the destination and automatically generates a three-dimensional flight route by reflecting the altitude change value.

4B is a process in which the unmanned aerial vehicle 100 flies and acquires a low altitude crop image through the image acquisition unit 130. At this time, the flight moves along the flight path obtained in (a) of FIG. 4.

The unmanned aerial vehicle 100 moves along the flight path and acquires a low-altitude crop image of about 20 meters, and transmits the acquired low-altitude crop image to the control center 200 in real time.

4C is a process of reading a crop in the control center 200 receiving a low altitude crop image and extracting the coordinates of the crop. At this time, crop reading can be performed through artificial intelligence (AI), but is not limited thereto.

4D is a process of producing an orthodox map and a theme map in the control center 200.

The orthogonal map is a map made to be the same scale by correcting the distortion of each point on the low-altitude crop image caused by remarks on the ground surface. Specifically, the orthogonal map is a diagram obtained by correcting the geometric distortion of the low-altitude crop image obtained from the low-altitude unmanned aerial vehicle 100 due to the undulations of the terrain and transforming it into a shape when all objects are viewed vertically, including coordinates and periods.

Thematic map is a map created for the purpose of expressing a specific subject. According to an embodiment of the present invention, the subject map may be a map in which crops are divided and displayed according to parcels, but is not limited thereto.

The crop cultivation information acquisition system 10 reads the type of crop at the same time as photographing through the low-altitude unmanned aerial vehicle 100, displays it on a map, and derives the cultivation area. At this time, the crop cultivation information acquisition system 10 can complete shooting and result generation at the same time by providing a real-time cultivation area, and it is possible to check the cultivation fluctuations in time series.

The low-altitude unmanned aerial vehicle 100 may perform a field survey, set a take-off and landing point, receive a flight path, photograph a normal and low altitude, and transmit the acquired data.

The control center 200 may set GIS data, generate an optimal flight path, transmit a flight path, and receive and manage data. In addition, the control center 200 analyzes low-altitude crop images to classify crops, calculates cultivation area, secures subject map GUI, AI learning data, inputs GIS spatial information, and manages data servers. have.

Here, the GUI refers to an environment in which a user can work through graphics when exchanging information with a computer.

Accordingly, the crop cultivation information acquisition system 10 grasps the crop cultivation status and growth status through low-altitude 3D terrain flight by reflecting the characteristics of the mountainous terrain survey. Real-time image data sent from the field can be used to create a subject status map on the day of shooting through image analysis, which is a technology that matches the goal of agricultural observation, and through real-time image transmission and crop analysis technology, the timeliness and accuracy of observation work can be secured. I can.

According to an embodiment of the present invention, the low altitude unmanned aerial vehicle 100 and the control center 200 exchange data in real time, and images can be immediately transmitted using 4G, 5G wireless networks and Wifi, and data deletion transmission function Can be done.

The low-altitude unmanned aerial vehicle 100 may transmit the obtained low-altitude crop image to the control center 200 in real time, and additionally transmit latitude, longitude, state information, and gimbal state information.

5 is a conceptual diagram illustrating a relationship between an altitude and a photographing area of a low-altitude unmanned aerial vehicle according to an embodiment of the present invention.

Conventionally, after orthogonal correction of an image acquired by a drone from 100 to 150 m in the air, clinical information of the cultivation site was obtained. In the case of the present invention, an image may be taken in a section of 10 to 30 m, and in particular, clinical information of a cultivation area may be acquired using an image acquired in a low altitude flight of 20 m.

According to a conventional drone, it is possible to read an image at a resolution of about 4 to 5 cm from 150 m above the air, and it is practically difficult to grasp the type of crop at that resolution. According to the drone of the present invention, since the image is read at a resolution of 0.5 cm in the air above 20 m, it is possible to predict the type of crop.

The low-altitude unmanned aerial vehicle 100 can accurately classify crops by forming a sufficient resolution by capturing an image in a range of 10 to 30 m.

Referring to FIG. 5, while a conventional drone can capture an area of 240 m width at an altitude of 150 m, the drone of the present invention can capture an area of 32 m width at an altitude of 20 m. Accordingly, the area is 56 times different, and due to the high resolution difference, crops can be classified only with images taken by the drone of the present invention, and the growing state can be confirmed.

The low-altitude unmanned aerial vehicle 100 is capable of precise photographing at a low altitude, avoiding hazards, shortening working time, and flying in the same trajectory. In addition, the transmission of the flight path has the effect of improving the flight work efficiency of the low-altitude unmanned aerial vehicle 100, enabling the route change in the low-altitude unmanned aerial vehicle 100 and the control center 200, and managing the flight record. .

In the case of a conventional shooting from 150 meters above, the 4 cm resolution is available, but it is impossible to distinguish the crops. According to the present invention, since the shooting is taken at 20 m, it is possible to distinguish the crops. Spatial information of each lot, particularly location information and boundary information, can be easily obtained from a database (GIS) that manages farmland information. The flying height of the drone is preferably 10 to 30 m, especially 15 to 25 m, and the upper height is to avoid collisions with, for example, telephone poles or power lines, and when driving higher than the upper height, it is considered difficult to recognize crops. .

In addition, the low-altitude unmanned aerial vehicle 100 performs low-flying in preparation for meteorological factors such as fog, gusts, low pressure, etc., changes the execution schedule, and includes a cooling device in preparation for high-temperature damage. I can.

6 is a view showing a flight path of a low-altitude unmanned aerial vehicle according to an embodiment of the present invention.

The low-altitude unmanned aerial vehicle 100 may set take-off and landing points 102 and 104. Here, the take-off and landing points 102 and 104 are points where both take-off and landing are possible.

The take-off and landing points 102 and 104 are easily accessible, and can be set to a place where there are no telephone poles or arboretum and easy to secure visibility. In addition, take-off and landing points can be set in a place that does not interfere with the traffic of resident vehicles.

According to an embodiment of the present invention, the take-off and landing points 102 and 104 are set to two, but are not limited thereto, and at least one take-off and landing point may be set.

Referring to FIG. 6, the low-altitude unmanned aerial vehicle 100 may obtain a low-altitude crop image while flying in a plurality of sub-areas. In this case, the sub-region may be divided, which may be divided into at least two regions based on an area, an elevation difference, and the like.

The take-off and landing points 102 and 104 may be set by the low-altitude unmanned aerial vehicle 100 or set at the control center 200, and when the low-altitude unmanned aerial vehicle 100 is set, the point coordinates can be transmitted to the control center 200. have.

7 is a view in which a representative lot of parcels is extracted according to an embodiment of the present invention, and a flight trajectory is generated according to the extracted representative parcels.

Lot information about the lot can be obtained from the Farmland Information Database (GIS). The lot information includes divisional boundary information between parcels and elevation information of areas included in the parcels. In this embodiment, the representative lot of the lot may be a central point of each lot, or a feature point that shows the characteristics of the lot.

Referring to FIG. 7, a path connecting the representative points of the parcel may be represented by marking the center point of each parcel as a representative parcel point.

The path following the representative point of FIG. 7 may be an actual driving path of the low-altitude unmanned aerial vehicle 100, but is not limited thereto.

The flight area on the left has a total flight distance of 3.83 km, an estimated flight time of 12 minutes and 23 seconds, and a total of 73 points. The flight area on the right has a total flight distance of 6.21Km, an estimated flight time of 17 minutes 12 seconds, and a total of 94 points. Here, the drones in the flight area on the right side and the right side have the same flight altitude as 20m.

Accordingly, the driving route of the unmanned aerial vehicle 100 may be determined to minimize the sum of connection distances connecting representative parcels. The driving route may be predetermined by a processor inside the drone or a processor located in the server, and the route may be changed while driving.

8 is a diagram illustrating another example of processing parcel image information according to an embodiment of the present invention.

A1 to a8 represent parcel representative points representing each parcel. In the case of a8-1 and a8-2, it is an example of processing one lot but two representative points. If the parcel is elongated and the center is narrow, the center of each area can be determined as the parcel representative point in a situation where the center of the parcel can be distinguished into two areas rather than the parcel representative point.

In addition, the representative lot of the lot may be a central point as well as a feature point that shows the characteristics of the lot. For example, the parcel image may include an image of a road and a farm road indicating the boundary of the parcel. The parcel image is irrelevant to the crop, and in order to more accurately recognize the crop of the parcel, it is necessary to determine a point that represents the crop characteristics of the parcel. For example, the image is analyzed in frequency domain for each sub-block It is also possible to adopt the position of a sub-block having a frequency domain value as a feature point.

h1 to h8-2 shows the altitude of each parcel center point. In determining the movement path of the drone, it is preferable that the distance between the center points of each parcel is not a distance on a plan view, but a distance on a three-dimensional spatial coordinate. This is because it is possible to obtain an image with a resolution that can distinguish crops and avoid collisions with the ground and obstacles.

9 is a diagram illustrating a concept of performing image learning through AI according to an embodiment of the present invention, and recognizing the type of crop through this.

The artificial neural network can perform distinct operations of a learning step and an application step. In the learning step, the artificial neural network receives training data, generates a feature value according to a predetermined weight value, and classifies the crop type using the generated feature value.

The artificial neural network can be implemented in the form of software executed by a processor. It is preferable that the training data be provided with various images for each season so as to cover various crops. Training data may already include information on crops (type of crop, season information, crop stage information, etc.), and supervised learning is possible using this information.

Through learning through training data, a weight value included in the artificial neural network may be determined within a range that minimizes errors in crop recognition. The artificial neural network may be, for example, a convolutional neural network. In the case of a convolutional neural network, it is advantageous for image processing and object recognition.

10 is a view showing a state of a crop by using an image acquired through a low-altitude drone flight according to an embodiment of the present invention.

Figure 10 (a) is a view showing the progress of the harvest, Figure 10 (b) is a view showing the cabbage soft rot, Figure 10 (c) is a view showing normal growth, Figure 10 (d) is It is a figure showing the lack of lime, and FIG. 10(e) is a figure showing high temperature damage.

The crop cultivation information acquisition system 10 may diagnose the degree of harvesting progress and the situation of various diseases and pests through the low-altitude flight of the low-altitude unmanned aerial vehicle 100. At this time, it is possible to obtain crop cultivation information through low altitude flight.

The crop cultivation information may include the type of crop and the state of the crop. The state of the crop includes the nature, size, and color of the crop, and through image analysis, it is possible to determine the crop's cultivation stage and the possibility of diseases and pests. In addition, information on the harvesting progress of crops can be additionally confirmed through image analysis.

According to an embodiment of the present invention, the state of a crop according to cabbage may be determined to be in good condition through cabbage softness, lime deficiency, high temperature damage, etc., and the cause may be analyzed and prevented. Specifically, it is possible to determine the type of crop and the state of the crop through the low altitude crop image acquired through the low altitude unmanned aerial vehicle 100.

The accumulated growth data can be accumulated as AI learning data to read crop reading, growth, and damage from pests.

According to another embodiment of the present invention, the crop cultivation information acquisition system 10 performs spectroscopic photographing to obtain a vegetation index, and measures ¼ of the body weight at ½ of the growth period, and the measurement item after harvesting is the height, The vegetation index and correlation analysis with the growth data can be performed by measuring the plant height and fresh weight of the final crop after completion of growth, and comparing the NDVI acquired by photographing with the actual crop growth data.

The vegetation index is based on the phenomenon that the spectral reflection characteristic of chlorophyll, which is directly proportional to the amount of nitrogen in the plant, mainly absorbs red and blue wavelengths, while reflecting near infrared (NIR) wavelengths. It is an index to evaluate the relative distribution and activity of green plants by combining them. The Normalized Difference Vegetation Index (NDVI) calculated using the reflectance of the red wavelength and near-infrared wavelength can be calculated. When shooting RED-edge and Sequoir, blur due to vibration and speed may occur, so make a work plan so that the shooting can proceed when the illumination is secured, and the camera sensitivity is set to ISO 400 or higher, and a shutter speed of 1/250 or higher is secured. can do. Shooting on a bright, cloudy day may be desirable to avoid reflectance distortion caused by direct light.

11 is a diagram illustrating extraction of an orthogonal image through low-altitude terrain flight according to an embodiment of the present invention.

FIG. 11A shows Fam Map lot information, and FIG. 11B shows DSM altitude information. FIG. 11(c) may show a flight area created based on FIGS. 11(a) and (b).

(D) of FIG. 11 is a diagram of obtaining a low-altitude crop image through terrain flight, (e) of FIG. 11 shows the acquired orthogonal image, and (f) of FIG. 11 is an image that can identify crops from the acquired orthogonal image. Represents.

The crop cultivation information acquisition system 10 can distinguish crops from the seedling period through a low-altitude mission flight within 25 meters, and the process of identifying cultivated crops can be faster and more accurate. In addition, the collected growth data can be accumulated as AI learning data, such as crop reading and growth, and damage data from pests.

The drone of the present invention can transmit an image acquired through a camera through a communication unit to an external server in real time. The server can analyze the type and distribution of crops quickly or in real time through image processing on the acquired images.

In addition, if the existing crop information is further included, the flight of the drone may be modified in consideration of the existing crop information. For example, in the case of onions, vertical images cannot be distinguished, and it is necessary to shoot images at 45 degrees from the side. Therefore, in determining the path of the drone, it is desirable to acquire images at an oblique angle at a height lower than 20 meters, for example, in a section of 10 to 15 meters, in the case of a parcel with a history of onion cultivation, using the history information of the crop. Do. In this case, it is preferable that the location of the drone is not the representative point of the lot, but the point that the drone's camera looks at is the representative point of the lot.

12 to 14 are flowcharts illustrating a process performed by the crop cultivation information acquisition system to acquire crop cultivation information according to an embodiment of the present invention. The process performed by the crop cultivation information acquisition system to acquire crop cultivation information is performed by the crop cultivation information acquisition system, and detailed descriptions and overlapping descriptions of operations performed by the crop cultivation information acquisition system will be omitted.

12 is a flowchart illustrating a process for producing a subject map in a control center according to an embodiment of the present invention.

The subject map production in the control center may be performed by receiving a low-altitude crop image acquired from a low-altitude unmanned aerial vehicle (S1210).

The low altitude crop image may be transmitted to the data server via the NAS server (S1220) (S122). Here, the NAS server represents a file-level computer storage device connected to a computer network.

The low-altitude crop image is transmitted to the AI server to perform crop reading (S1230). In addition, the low-altitude crop image may be additionally transmitted to the image reader to perform crop reading (S1232).

Step S1230 may crawl image sources such as collection of photographs of crops observed on public websites of the Rural Economic Research Institute or Rural Development Administration, acquisition of drone images of agricultural land, and due diligence when cultivating off-side. Crawling is a technology that collects documents distributed and stored in a myriad of computers and includes them as an index of search targets, and how quickly a certain kind of technology is included in the search target is a factor that determines its superiority.

In step S1230, the image source may be crawled and the crop may be read through a reading algorithm. Reading algorithms can use Vggnet, Alexnet, Squeeznet, Resnet, Densenet, and the like.

In this case, accuracy may be improved through cross-validation of the read crops according to the crop readings performed in steps S1230 and S1232.

The crops read in step S1230 and step S1232 may classify the same images through image classification (S1240).

Further, in step S1250, the orthogonal correction process may be omitted for the low altitude crop image.

When the image classification (S1240) and orthogonal correction (S1250) are completed, a theme map may be produced (S1260). The produced theme map is stored in the data server (S1270).

13 is a flowchart illustrating a process performed by the AI server in step S1230 of FIG. 12 according to an embodiment of the present invention.

The AI server may acquire an image (S1310). Here, the image is a low-altitude crop image acquired from a low-altitude unmanned aerial vehicle.

When the AI server acquires the image, it may extract the image (S1320). In the image extraction in step S1320, crop images are extracted as much as a preset area from the low-altitude crop images acquired in each lot as shown in FIG. 13. For example, the image extraction is an image located in the middle area of the acquired image, and may be extracted as a rectangular area.

In step S1330, the AI server may classify the crops through artificial intelligence by extracting the crop images from the low-altitude crop images acquired in each parcel. Classification of crops through artificial intelligence can be confirmed through FIG. 9, and each crop can be classified according to type.

In step S1330, the crop is not classified, and the process returns to the step of acquiring the image (S1310). When the crop classification is completed, the coordinates of the image may be extracted in step S1340.

In step S1340, the image coordinate extraction can extract the coordinates according to each lot through the GIS through the obtained low-altitude crop image, or take a low-altitude crop image at each lot through the GPS attached to the low-altitude unmanned aerial vehicle. The coordinates can be extracted through the location of the unmanned aerial vehicle in the case where

When the coordinate extraction is completed, in step S1350, each area may be calculated. Areas are calculated to generate crop maps, and can be checked through GIS based on each parcel.

According to an embodiment of the present invention, data for each growth stage for AI learning can be identified as a seedling period, early growth period, maturity period, harvest period, and the like, and when data is accumulated, growth, pests, and the like can be additionally read.

14 is a flowchart illustrating a web-based real-time time series value drawing according to an embodiment of the present invention.

Deriving a web-based real-time time series value picture is a step of creating an orthogonal image (S1410), classifying a crop through a photographed image (S1420), producing a cultivation map for each crop and lot (S1430), cultivation of cabbage Separating the lot (S1440), displaying a change in the lot of cabbage cultivation (S1450), and a step of time-series (S1460).

The web-based real-time time-series value plot can display crop cultivation status, shipping area, distribution map at planting time, distribution map at harvest time, etc. all as figures and values.

In the step of creating an orthogonal image (S1410), an orthogonal image may be generated based on a low-altitude crop image acquired from a low-altitude unmanned aerial vehicle.

In the step of classifying the crop through the photographed image (S1420), the crop may be classified through the AI server of FIG. 14.

In the step of producing a cultivation map for each crop and lot (S1430), the crops classified in step S1420 may be displayed in different colors to display crops according to each lot to produce a cultivation map.

In the step of separating the cabbage cultivation lot (S1440), only the cabbage may be extracted from the cultivation map obtained in step S1430. Step S1440 is shown as an embodiment, it can be separated into other crop cultivation lot other than cabbage.

In the step of displaying the change of the cabbage cultivation lot (S1450), the change may be displayed by comparing the previously identified cabbage cultivation lot with the current low-altitude unmanned aerial vehicle. Specifically, when the number of lots of cabbage cultivation increases, areas different from the existing areas can be displayed in different colors to confirm the added lot.

In the step of time-series conversion (S1460), an area in which the change of the cabbage cultivation lot is displayed may be displayed on an orthogonal map to form the area. Specifically, in the step of time-seriesizing (S1460), observation results are arranged in a series according to a certain standard.

The web-based real-time time series value plot is a time series of changes in the area of cultivation in real time using GIS technology and data analysis image realization technology.

Crop cultivation information acquisition system 10 by crop item. The cultivation area, planting area, and shipping area for each lot can be calculated as images and values, and crop classification, planting season distribution, and harvest season distribution can be implemented using database analysis tools. At this time, low-altitude photos of the site are arranged on each lot, and the site photos can be opened with a click.

In FIGS. 12 to 14, it is interposed that each process is sequentially executed, but this is only illustrative, and those skilled in the art are shown in FIGS. 12 to 14 within the range not departing from the essential characteristics of the embodiment of the present invention. By changing the described order, executing one or more processes in parallel, or adding other processes, various modifications and variations may be applied.

Therefore, the crop cultivation information acquisition system 10 transmits images of crops photographed by drones, transmits images of drones photographed to a specific folder on the server, generates artificial intelligence deep learning reading models, operates reading models, and compares each result value. Python code to read crops, a module that stores and manages the read values in a database, a module that displays the reading results stored in the database on a map, a module that displays a picture of the crop and the read results when a location on the map is selected, etc. It may be configured to perform the above-described processes, but is not limited thereto.

15 is a block diagram illustrating a learning concept for crop analysis performed in an AI server according to an embodiment of the present invention.

A generative model (GAN) is a machine learning that automatically creates an image that is close to the real world while the generative model and the discriminant model compete.

15 shows a learning concept performed by a processor of an AI server as a kind of hostile neural network. The processes performed by the hostile neural network shown in FIG. 15 are a series of programs for executing the hostile neural network, and are stored in the memory of the AI server, and include the following operations performed by the processor of the AI server.

First, the input information 1610 for learning of the hostile neural network includes a plurality of training images obtained from a corresponding lot. As additional input information 1610, time information (climate information, shooting time) when the image was acquired, location information (location information on GPS, identification information of the parcel, etc.), color information (image acquired at the parcel) The main color information in) is further included as label information (which is also training data).

As shown in FIG. 15, the generator 1630 generates a fake image by inputting the label information 1610 and the random noise 1620. The classifier 1640 is composed of a plurality of convolution filters, an activation layer, and a pooling layer, and determines Real/Fake using the final feature values.

The update of the filter coefficients of the classifier 1640 and the generator 1630 is a feedback on the fake image generated through the generator 1630 and the classification result (Real/Fake response) for the training image 1650, which is a real sample. Through this, the filter coefficient is updated in a direction in which a more accurate classification result can be obtained. Through this learning process, the filter coefficient of the generator 1630 may be determined. From a physical point of view, the classifier 1640 and the generator 1630 substantially correspond to a processor, and may be understood as terms representing detailed operations performed by the processor.

16 is a conceptual diagram illustrating a method of generating an input image for image classification (S1240) through the generator 1630 determined through FIG. 15.

If the image acquired through the current drone can be classified with a sufficient level of confidence, the need to perform this process is small. However, when it is difficult to specify the type of crop with only the currently acquired image, it is desirable to determine the type of crop using a composite image generated through a hostile neural network.

In FIG. 16, the generator 1630 refers to the generator 1630 for which learning is completed through FIG. 15. Here, the inputs are the current acquired image 1710 and a label 1610 for the acquired image 1710. The information 1610 included in the label may include time information, location information, and color information of the time when the image is captured. Inputs for image classification (S1240) include a synthesized image 1660 synthesized by the creator 1630 by inputting label information 1610 as an input, and a currently acquired image 1710. If the feature value extracted through the synthesis image 1660 and the feature value extracted through the acquired image 1710 are summed according to a predetermined weight, and then image classification is performed using the summed feature value, When compared to the case of performing classification, the reliability of classification can be further increased.

Even if all the components constituting the embodiments of the present invention described above are described as being combined into one or operating in combination, the present invention is not necessarily limited to these embodiments. That is, within the scope of the object of the present invention, all the constituent elements may be selectively combined and operated in one or more.

The above description is merely illustrative of the technical idea of the present invention, and those of ordinary skill in the technical field to which the present invention belongs can make various modifications, changes, and substitutions within the scope not departing from the essential characteristics of the present invention. will be. Accordingly, the embodiments disclosed in the present invention and the accompanying drawings are not intended to limit the technical idea of the present invention, but are for description, and the scope of the technical idea of the present invention is not limited by these embodiments and the accompanying drawings. . The scope of protection of the present invention should be interpreted by the following claims, and all technical ideas within the scope equivalent thereto should be construed as being included in the scope of the present invention.

10: Crop cultivation information acquisition system
100: low altitude unmanned aerial vehicle
200: control center

Claims (12)

A control center for generating a flight path according to a parcel formed in consideration of preset terrain information and take-off and landing points; And
It includes a low-altitude unmanned aerial vehicle that performs low-altitude flight according to the flight path and obtains environmental information or low-altitude crop images,
The control center analyzes the low altitude crop image to obtain crop information, displays the crop information according to coordinates, and generates a crop map,
The low altitude unmanned aerial vehicle,
A driving unit that generates a driving force that enables the flight of the low-altitude unmanned aerial vehicle;
An image acquisition unit that moves along the flight path and obtains the low-altitude crop image; And
And a processor for controlling the driving unit and the image acquisition unit, and resetting the flight path in consideration of the environment information and the obtained low altitude crop image,
The processor,
First crop recognition is performed by analyzing the first low altitude crop image acquired by the image acquisition unit set to face directly downward,
When crop information is not obtained through the primary crop recognition, the second crop image obtained by adjusting the angle of the image acquisition unit set to face the direct downward direction is received and analyzed to perform secondary crop recognition. Perform,
When crop information is not obtained through the secondary crop recognition, the first low-altitude crop image and the second low-altitude crop image are synthesized through a hostile generated neural network (GAN) to generate a synthetic image, and the generated synthetic image Crop cultivation information acquisition system, characterized in that performing the third crop recognition by analyzing the. The method of claim 1,
The low altitude unmanned aerial vehicle,
A sensor unit for acquiring the surrounding environment information; And
Transmitting the obtained environmental information, the low altitude crop image or the crop information to the control center, crop cultivation information acquisition system further comprising a communication unit receiving the flight path. The method of claim 2,
The sensor unit,
A front sensor detecting an obstacle positioned in front;
A lower sensor for detecting an obstacle positioned below;
An ultrasonic sensor that detects surrounding obstacles through ultrasonic waves; And
Including a proximity sensor for detecting an obstacle in proximity,
The sensor unit includes at least one obstacle information detected by the front sensor, the lower sensor, the ultrasonic sensor, and the proximity sensor to obtain the environment information. The method of claim 1,
The processor,
When an obstacle detected according to the acquired environmental information is located in the flight path, a movement path for moving to the next point is re-analyzed in consideration of the number and size of the detected obstacles, and the flight path is reset to establish a reset path. Generating and transmitting the reset route to the control center, crop cultivation information acquisition system, characterized in that. delete The method of claim 1,
The tertiary crop recognition,
The first low-altitude crop image or the second low-altitude crop image and the label information of the first low-altitude crop image or the second low-altitude crop image are input and synthesized through a generator through a hostile generation neural network (GAN). It is performed by analyzing the composite image,
The label information includes time information, location information, and color information when the first low-altitude crop image or the second low-altitude crop image is acquired. The method of claim 1,
The control center,
A route setting unit extracting a representative point of the parcel based on the preset topographic information including parcel information, and setting a flight path of the low-altitude unmanned aerial vehicle in consideration of the representative point of the parcel; And
Crop cultivation information acquisition system comprising a crop detection unit for generating a crop map by analyzing the low altitude crop image obtained according to the flight path. The method of claim 7,
The crop detection unit,
An image analysis unit for generating the crop information by analyzing the low altitude crop image to classify the crop;
An area calculation unit for calculating a cultivation area according to the crop information among the total area according to the flight path;
A coordinate extraction unit for extracting coordinates of the crop information; And
A system for obtaining crop cultivation information including a map generator for generating a crop map using the flight path, the crop information, the cultivation area and the coordinates. The method of claim 7,
The path setting unit,
Generates the flight path in consideration of the flight trajectory according to the extracted representative point representing the center point of the parcel,
The lot information is a system for obtaining crop cultivation information, characterized in that it includes information on a boundary between divisions and height information of areas included in the lot. A driving unit that generates a driving force that enables flight along a flight path generated in consideration of preset terrain information;
A sensor unit for acquiring surrounding environment information;
An image acquisition unit that performs low altitude flight according to the flight path and acquires a low altitude crop image;
A processor that analyzes the low-altitude crop image to obtain crop information, and displays the crop information according to coordinates to generate a crop map; And
A communication unit for transmitting the obtained low-altitude image or an operation result in the processor to an external communication capable device,
The processor,
First crop recognition is performed by analyzing the first low-altitude crop image acquired by the image acquisition unit set to face directly downward,
When crop information is not obtained through the primary crop recognition, the second crop image obtained by adjusting the angle of the image acquisition unit set to face the direct downward direction is received and analyzed to perform secondary crop recognition Perform,
When crop information cannot be obtained through the secondary crop recognition, the first low-altitude crop image and the second low-altitude crop image are synthesized through a hostile generated neural network (GAN) to generate a synthetic image, and the generated synthetic image Low-altitude unmanned aerial vehicle, characterized in that to perform the third crop recognition by analyzing. The method of claim 10,
The processor,
When an obstacle detected according to the acquired environmental information is located in the flight path, a movement path for moving to the next point is re-analyzed in consideration of the number and size of the detected obstacles, and the flight path is reset to establish a reset path. A low-altitude unmanned aerial vehicle, characterized in that to generate and transmit the reset route to the control center. delete

Images