JP2018068221A - Unmanned device for harmful animal extermination, harmful animal extermination method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a harmful animal extermination device for exterminating a harmful animal intruding into a farm and a ranch.SOLUTION: An unmanned device 1 for harmful animal extermination includes autonomous moving means that moves autonomously, image acquisition means for acquiring an image of an intrusion prohibition area, determination means for recognizing whether or not an object in the image is an extermination target, and determining whether or not the extermination target has intruded into the intrusion prohibition area, and extermination means for exterminating the extermination target.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a pest control drone, a pest control method, and a program.

  Conventionally, invasion prevention fences have been proposed in order to protect crops from pests such as deer and wild boar.

JP 2009-11280 A

  By the way, the intrusion prevention fence is effective for preventing the invading pests from entering the farm or ranch, but cannot intrude the invading pests.

  The present invention is an attempt to solve such a problem, and an object of the present invention is to provide a pest-control drone that controls a pest that has invaded a farm or ranch.

The first invention recognizes whether or not an autonomous moving means for autonomously moving, an image acquiring means for acquiring an image in an intrusion prohibited area, and an object in the image is a target for removal, Determination means for determining whether or not the removal target has entered; a removal means for removing the removal target;
An unmanned aerial vehicle for pest control.

  According to the configuration of the first aspect of the invention, the pest control drone can control the control target that has entered the intrusion prohibited area while autonomously moving. That is, pests that have invaded farms and ranches can be eliminated.

  According to a second invention, in the configuration of the first invention, the pest control means includes a capture means for capturing the control target and / or an exclusion means for removing the control target from the intrusion prohibited area. It is a drone for use.

  According to the structure of 2nd invention, the pest control apparatus can implement capture and / or exclusion.

  According to a third invention, in the configuration of the second invention, the exclusion method performed by the exclusion means includes sound generation, light irradiation, odor generation, movement in a specific movement mode, or a combination thereof. An unmanned aerial vehicle for pest control.

  According to the configuration of the third invention, an appropriate exclusion method can be implemented from among a plurality of exclusion methods.

  According to a fourth aspect of the present invention, in the configuration of the third aspect, the exclusion method is defined for each type of the target to be controlled, and the drone controlling drone tracks the target to be controlled while tracking the target to be controlled. A pest control drone that operates the control means according to the type.

  According to the configuration of the fourth aspect of the invention, it is possible to perform appropriate removal according to the type of the removal target while tracking the removal target.

In a fifth aspect of the invention according to any one of the third and fourth aspects, the exclusion method is defined by distinguishing between a daytime exclusion method and a nighttime exclusion method,
The exclusion means is a pest-control drone configured to selectively use the daytime exclusion method and the nighttime exclusion method.

  According to the configuration of the fifth aspect of the invention, an appropriate exclusion method can be implemented according to daytime or nighttime.

  According to a sixth invention, in the configuration of any one of the first to fifth inventions, the image acquisition means is configured to acquire an image in visible light or an image in infrared according to the intensity of ambient light. It is a drone for pest control.

  According to the configuration of the sixth aspect of the invention, it is possible to acquire an image for determining whether or not it is a target for removal according to the intensity of ambient light.

  According to a seventh invention, in the configuration of any one of the first to sixth inventions, the determination means uses deep learning to recognize whether or not the object in the image is the removal target. It is an unmanned aerial vehicle for pest control that refers to the generated data.

  According to the configuration of the seventh aspect of the invention, it is possible to accurately recognize whether or not it is a target for removal based on data generated by deep learning.

  According to an eighth aspect of the present invention, in the configuration of any one of the second to seventh aspects, based on the image acquired by the image acquisition means, it is determined whether or not the capture of the removal target has been successful, If it is determined that it has not succeeded, it calculates the success probability when the capture is performed again, and if the success probability is equal to or greater than a predetermined value, the removal means performs the capture of the removal target again. If the success probability is less than the predetermined value, the pest control drone implements exclusion of the control target.

The ninth invention is a pest control method implemented by an autonomously movable pest control drone, an image acquisition step of acquiring an image in an intrusion prohibited area, and whether an object in the image is a target to be controlled A pest control method comprising: a determination step of recognizing whether or not the control object has entered the intrusion prohibited area; and a control step of controlling the control target.

  A tenth aspect of the invention is a computer that controls an autonomously movable pest control drone, an image acquisition unit that acquires an image in an intrusion prohibited area, recognizes whether an object in the image is a target for control, A program for functioning as a determination unit that determines whether or not the removal target has entered the intrusion prohibited area and a removal unit that removes the removal target.

  ADVANTAGE OF THE INVENTION According to this invention, the pest which invaded the farm and the ranch can be exterminated.

It is the schematic which shows the effect | action of the unmanned aircraft for pest control which concerns on embodiment of this invention. It is the schematic which shows the effect | action of the drone for pest extermination. It is the schematic which shows the effect | action of the drone for pest extermination. It is the schematic which shows the effect | action of the drone for pest extermination. It is a figure which shows the structure of the drone for pest extermination. It is the schematic which shows a capture net. It is a figure which shows the function structure of the drone for pest extermination. It is a figure which shows the characteristic database. It is a figure which shows the exclusion method database. It is a figure which shows the exclusion method combination database. It is a schematic flowchart which shows operation | movement of the unmanned aircraft for pest extermination. It is a schematic flowchart which shows operation | movement of the unmanned aircraft for pest extermination. It is a figure which shows the structure of the drone for pest extermination. It is a figure which shows the characteristic database. It is a figure which shows the exclusion method combination database. It is a schematic flowchart which shows operation | movement of the unmanned aircraft for pest extermination.

  Hereinafter, modes for carrying out the present invention (hereinafter referred to as “embodiments”) will be described in detail. In the following description, the same reference numerals are given to the same components, and the description thereof is omitted or simplified. Note that description of configurations that can be appropriately implemented by those skilled in the art will be omitted, and only the basic configuration of the present invention will be described.

<First embodiment>
FIG. 1 to FIG. 4 are schematic views showing the operation of the unmanned aerial vehicle for pest control (hereinafter referred to as “unmanned aircraft”) according to an embodiment of the present invention. A drone 1 shown in FIG. 1 and the like is an example of a drone extermination drone. The drone 1 is an unmanned vehicle that travels on the ground. The drone 1 is disposed in the area F. Area F is an example of an intrusion prohibited area. The drone 1 is configured to be able to communicate with the base station 100, and appropriately communicates with the base station 100 and follows instructions from the base station 100. The moving area of the drone 1 is limited so that it can move only between the base station 100 and the area F and within the area F. The drone 1 is configured to charge or replace the battery at the base station 100 when the remaining battery level is equal to or lower than a predetermined reference value. The drone 1 is controlled so as not to approach another object within a predetermined distance, and is configured to avoid a collision with another object.

  1 to 4, the region F includes small regions F1, F2, F3, and F4. The region F is a region that is slightly wider than the region where the small regions F1, F2, F3, and F4 are combined.

  The drone 1 is configured to exterminate the harmful animals from the area F when the harmful animals such as the deer 301 and the wild boar 303 enter the area F. A pest is an example of an extermination target. The small area F1 is a farm, for example.

  As shown in FIG. 2, for example, when the deer 301 enters the area F, the drone 1 captures the deer 301 (see FIG. 3) or excludes the deer 301 from the area F (takes out). (See FIG. 4). In this specification, the term “disinfect” is used to include capture and exclusion.

  Since the area F is an area having a larger area than the small area F1 or the like, not only the harmful animals that have entered the small area F1 and the like are excluded, but also when the pest elimination is performed quickly, the small area F1 and the like. Can be prevented from entering. Note that the area F1 and the like are not limited to farms and may be ranches, for example, as long as they are areas that require protection from harmful animals.

  Hereinafter, the configuration of the drone 1 will be described.

As shown in FIG. 5, the drone 1 has a housing 10. Arranged inside the housing 10 are a computer that controls each part of the drone 1, a wireless communication device, a positioning device using GPS (Global Positioning System), an acceleration sensor, a gyro sensor, a magnetic sensor, a battery, and the like. Four wheels 12 (two front wheels and two rear wheels) are arranged in the housing 10. A motor (not shown) is disposed inside the housing 10, and two rear wheels of the four wheels 12 are connected to the shaft of the motor, so that the vehicle can travel vertically and horizontally. . The housing 10 is provided with a headlamp 14 and an antenna 16. The front of the drone 1 can be illuminated by the headlamp 14. Radio waves are transmitted and received from an antenna 16 constituting the wireless communication device, and communication with the base station 100 is possible. In the housing 10, a camera 20 is disposed on a shaft member 18 that can rotate and extend. The camera 20 is a visible light camera and can acquire an external image. At night, the drone 1 illuminates the front of the drone 1 with the headlamp 14 and acquires an image with the camera 20. In the housing 10, a light emitting device 24 and a sound generating device 26 are arranged on a shaft member 22 that can rotate and extend. In addition, a capture net launching device 28 is disposed in the housing 10. When the drone 1 determines that the object in the image acquired by the camera 20 is the deer 301 or the like to be removed, it launches the capture net launching device 28 while tracking the removal target by autonomous movement. The removal target is excluded from the region F by capturing the removal target or operating the light emitting device 24 and the sound generation device 26 in a predetermined manner. The light emitting device 24 and the sound generation device 26 are configured to exclude harmful pests, and the capture net launching device 28 is configured to capture the pests.

  The capture net launcher 28 is loaded with the capture net 30 shown in FIG. 6 in a folded state. When fired, the capture net 30 expands as shown in FIG. The capture net 30 is composed of a net body 30a having a small hole 30c in the center and a weight 30b arranged at the periphery of the net body 30a. For example, the mechanism described in Japanese Patent Application Laid-Open No. 2010-48539 can be used as a mechanism for launching the capture net 30. However, in the drone 1, the capture net 30 is fired by the control and mechanism of the drone 1 itself.

  FIG. 7 is a diagram illustrating a functional configuration of the drone 1. The drone 1 includes a CPU (Central Processing Unit) 50, a storage unit 52, a wireless communication unit 54, a GPS (Global Positioning System) unit 56, an inertial sensor unit 58, an image processing unit 60, a drive control unit 62, and a removal means control unit. 64, a distance sensor unit 66, and a power source unit 68.

  The drone 1 can communicate with the base station 100 by the wireless communication unit 54. The drone 1 receives various instructions from the base station 100 by the wireless communication unit 54 and transmits various information to the base station 100.

  The drone 1 can measure the position of the drone 1 itself by the GPS unit 56 and the inertial sensor unit 58. The GPS unit 56 basically receives radio waves from three or more GPS satellites and measures the position of the drone 1. The inertial sensor unit 58 measures the position of the drone 1 by accumulating the movement of the drone 1 from the starting point using, for example, an acceleration sensor and a gyro sensor. The drone 1 uses the position of the drone 1 itself for autonomous movement, measurement of the position to be removed, and the like.

  The image processing unit 60 allows the drone 1 to operate the camera 20 and acquire an external image.

  The drone 1 controls the movement of the four wheels 12 by the drive control unit 62.

  The removal means control unit 64 controls the capture net launching device 28, the light emitting device 24, and the sound generation device 26 to capture the harmful beasts to be removed, or to remove the harmful beasts from the region F. To implement.

  The distance sensor unit 66 measures the distance between the drone 1 and another object. The distance sensor unit 66 includes, for example, a laser rangefinder (laser range). The distance sensor unit 66 continuously measures the distance between the drone 1 and other objects.

  The power supply unit 68 is a replaceable rechargeable battery, for example, and supplies power to each unit of the drone 1.

  The storage unit 52 stores various data and programs necessary for autonomous movement, such as data indicating a movement plan for autonomous movement in the area F, which is a planned movement area, and data indicating the position and shape of topography and structures. Is stored. The storage unit 52 further stores the following programs and data.

  The storage unit 52 stores an autonomous movement program, an image acquisition program, a position information generation program, a determination program, an object recognition program, a removal program, a completion confirmation program, a movement range restriction program, and a collision avoidance program. The object recognition program is a part of the determination program, and is a general object recognition program that recognizes a category of an object in an image. Here, general object recognition means recognition of an object category. For example, general object recognition is to recognize whether an object in an image is a person, a deer, or a wild boar.

  The CPU 50 and the autonomous movement program are examples of autonomous movement means. The drone 1 moves autonomously, for example, refers to the terrain and the position of the structure in the region F stored in the storage unit 52 and circulates in the region F.

  The CPU 50 and the image acquisition program are examples of image acquisition means. The drone 1 operates the camera 20 to continuously acquire images in the area F.

  The CPU 50 and the position information generation program are examples of position information generation means. The position information is information indicating the position of the drone 1 itself and information indicating the position of the intruder. The drone 1 continuously generates and holds information indicating the absolute position and posture of the drone 1 by the GPS unit 56 and the inertial sensor unit 58. Further, by measuring the distance from a specific object at a plurality of different points by the distance sensor unit 66, for example, the deer 301 that has entered the small area F1 is positioned relative to the drone 1 or the small area F1. The relative position with the internal structure is calculated. Then, the drone 1 can calculate the absolute position of the deer 301 with reference to the absolute position of the drone 1 itself or the position of a structure or the like in the small area F1. Detailed description of position measurement is omitted.

  The CPU 50 and the determination program are examples of determination means. The determination program includes an object recognition program. The drone 1 refers to the feature data generated by the deep learning by the object recognition program. The feature data is stored in the storage unit 52. Deep learning is machine learning of a neural network having a multi-layer structure, and the field of image recognition is one of the promising fields of use.

  The drone 1 refers to the feature database shown in FIG. 8 by the object recognition program. The feature database is stored in the storage unit 52. The feature data stored in the feature database is data indicating a large number of features such as contours and directions of individual components with respect to the appearance of harmful beasts such as deer, wild boar, and monkey.

  The drone 1 can recognize the category of the object by comparing the feature of the object included in the image acquired by the camera 20 with the feature information of each category. Here, the category means the kind of pests such as deer, wild boar and monkey.

  Specifically, the drone 1 extracts a large number of features such as contours and directions of individual components from the image acquired by the camera 20. Then, the drone 1 determines the correlation (correlation degree) by comparing these features with the feature data of each category acquired by deep learning and stored in the storage unit 52. The higher the degree of correlation, the higher the possibility that an object in the acquired image belongs to a specific category. For example, when the degree of correlation is 0, the possibility of belonging to a specific category (hereinafter referred to as “category common probability”) is 0%, and when the degree of correlation shows the maximum value, the category common probability is 100. It is defined as%. The drone 1 determines that the category of the object in the acquired image belongs to a specific category when the category common probability is a predetermined reference value, for example, 95% or more.

  When the drone 1 recognizes that the object in the image belongs to the category of a specific pest by the object recognition program, the drone 1 determines that the object in the image is an object to be exterminated by the determination program. In the present embodiment, the target for extermination is a pest such as a deer, a wild boar, or a monkey.

  The CUP 50 and the removal program are examples of removal means, and are also examples of capture means and exclusion means. The drone 1 tracks a disinfection target by a disinfection program. The drone 1 controls the movement of the drone 1 so that an image to be removed is always included in an image acquired by the camera 20. That is, the drone 1 refers to the moving direction of the removal target in the image continuously acquired by the camera 20 and the movement speed of the removal target acquired by the distance sensor unit 66, and the removal target is captured in the image of the camera 20. To control the movement. Then, when the drone 1 can capture the removal target, or when receiving an instruction to capture the removal target from the base station 100, the drone 1 operates the capture net launching device 28 to remove the removal target. Perform subject capture. Specifically, the drone 1 calculates the relative position between the drone 1 and the removal target by the position information generation unit described above, calculates the vertical and horizontal angles at which the capture net 30 is fired, and removes it. It is configured to fire a capture net 30 against the subject. The accuracy of the relative position can be improved by setting the laser beam irradiation target of the laser rangefinder controlled by the distance sensor unit 66 as a target for removal and measuring the distance.

  If the drone 1 cannot capture the pest, does not receive an instruction to capture from the base station 100, or if the capture is not successful, the light emitting device 24 and / or the sound generator 26 is activated to eliminate the pests. The drone 1 refers to the exclusion method database shown in FIG. 9 by the removal program. The exclusion method database is stored in the storage unit 52. Data indicating a plurality of exclusion methods is stored in the exclusion method database. For example, as sound data, voice data such as deer warning sound (S1), gunshot (S2), elephant cry (S3), wolf cry (S4), etc. are stored in association with the identification numbers. ing. Similarly, a plurality of types of light, speed, and movement path are stored. Note that an odor generating device may be arranged in the drone 1, and an odor that a harmful animal dislikes may be added as an elimination method.

  The drone 1 refers to the exclusion method combination database (hereinafter referred to as “combination database”) shown in FIG. 10 by the removal program. The combination database is stored in the storage unit 52. The combination data stored in the combination database is data obtained by combining the exclusion methods shown in FIG. The combination data is data obtained by variously combining the exclusion methods of FIG. 9 such as a combination of deer warning sound (S1), speed 5 km / h (V1), and linear movement path (R3). The combination data includes an evaluation value indicating the effectiveness of the combination of exclusion methods. Evaluation values are defined for each type of pest. The drone 1 preferentially adopts a combination of exclusion methods having a high evaluation value for each type of pest.

  The feature database (FIG. 8), the exclusion method database (FIG. 9), and the combination database (FIG. 10) included in the drone 1 also include the base station 100. The base station 100 can update or change the data in these databases as necessary to rewrite the database of the drone 1.

  As described above, when the drone 1 determines that the object in the image acquired by the camera 20 is the object to be disinfected by the determination program, the drone 1 tracks the disinfection target based on the position information generated by the position information generation program. Upon start, the removal program is configured to capture or eliminate the removal object. That is, when the drone 1 determines that the intruder is a target for extermination, the drone 1 is effective based on the highly accurate position information generated by the drone 1 itself, and effective pest control according to the type of the target to be exterminated. Can be implemented.

  The CPU 50 and the completion determination program are examples of completion determination means and are part of the determination program. For example, if the drone 1 is determined to be within a predetermined range for a predetermined time in a state where the capture net 30 is entangled in the deer 301 that is a pest that is the target of extermination, the drone 1 determines that the capture of the extermination target has been completed. . The predetermined time is, for example, 2 minutes, and the predetermined range is, for example, within a radius of 5 m from the position of the removal target at the beginning of the predetermined time. The drone 1 makes the above determination based on the image acquired by the camera 20.

  On the other hand, if the drone 1 determines that the removal target is not located in the region F for a predetermined time, for example, the drone 1 determines that the removal of the removal target is completed. The predetermined time is, for example, 5 minutes. The drone 1 makes the above determination based on the image acquired by the camera 20.

  The CPU 50 and the movement range restriction program are examples of movement range restriction means. The drone 1 refers to the coordinate information of the area F stored in the storage unit 52 and the position information indicating the position of the drone 1 itself generated by the GPS unit 56 and the inertial sensor unit 58. It is determined whether or not the position is within the area F that is the area to be moved. If the position of the drone 1 is outside the region F, the drone 1 returns the drone 1 to the region F.

  The CPU 50 and the collision avoidance program are examples of collision avoidance means. When the distance between the drone 1 and another object (referred to as “object A”) measured by the distance sensor unit 66 is within a predetermined distance, the drone 1 is unmanned by the drive control unit 62. The distance between the machine 1 and the object A is made larger than a predetermined distance. Thereby, it is possible to avoid the drone 1 colliding with the object A. The predetermined distance is 2 m (meters), for example.

  Hereinafter, the operation of the drone 1 will be described with reference to FIGS. 11 and 12. The drone 1 circulates in the area F by autonomous movement, continuously acquires images with the camera 20, and performs image analysis (step ST1 in FIG. 11). When the drone 1 determines that the object in the image is a target for removal (step ST2), the drone 1 performs tracking and removal means operation (step ST3). When the drone 1 determines that the removal of the removal target has been completed (step ST4), it stops tracking and stops the removal means (step ST5). Then, the drone 1 transmits information indicating that the removal has been completed to the base station 100 (step ST6), and resumes patrol (step ST7).

  If the drone 1 determines that the condition for capturing the removal target is satisfied in the above-described step ST3 (step ST31 in FIG. 12), the position of the removal target is confirmed (step ST32), and the launch angle and launch of the capture net 30 are confirmed. The direction is calculated (step ST33), and the capture net 30 is fired (step ST34).

  When the drone 1 determines that the conditions for capturing are not satisfied (step ST31), the unmanned aircraft 1 selects a combination of exclusion methods with a high evaluation value according to the type of the harmful animal to be exterminated (step ST35). Exclusion by combination is performed (step ST36).

<Second Embodiment>
Next, a second embodiment will be described with reference to FIG. In addition, description is abbreviate | omitted about the structure which is common in 1st embodiment, and it demonstrates centering on a different part from 1st embodiment.

  As shown in FIG. 13, the drone 2 has a camera 21 in addition to the camera 20. The camera 20 is a visible light camera, and the camera 21 is a near infrared camera. The image processing unit 60 operates the visible light camera 20 or the near-infrared camera 21 based on the light intensity sensed by a light sensor (not shown) arranged in the drone 2. The optical sensor is composed of, for example, a photoresistor. The drone 2 activates the visible light camera 20 if the light intensity is at the daytime level, and activates the near-infrared camera 21 if the light intensity is less than the daytime level. Unlike the present embodiment, the functions of the visible light camera and the near-infrared camera may be realized by a hybrid camera that has both functions and is capable of switching both functions. Further, a far-infrared camera may be provided instead of or in addition to the near-infrared camera.

  As shown in FIG. 14, the feature database includes feature data under visible light and feature data under near infrared rays. The drone 2 refers to the visible light feature data when the image processing unit 60 operates the visible light camera 20, and refers to the near infrared feature data when the image processing unit 60 operates the near infrared camera 21. It is like that. Unlike the present embodiment, when the drone 2 has a far-infrared camera, the feature data is configured to include feature data under far-infrared rays.

  As shown in FIG. 15, in the exclusion method combination database, combinations of exclusion methods are defined separately for daytime and nighttime in addition to the types of harmful animals. For example, for the combination of S1, V1 and R3, the daytime evaluation value is “8” for deer but “5” for wild boar. The nighttime evaluation value of the combination is “9”, which is higher than the daytime evaluation value for deer. The drone 2 preferentially adopts a combination of exclusion methods having a high daytime evaluation value depending on the type of harmful animal during the daytime when the visible light camera 20 is operated. In addition, the drone 2 preferentially adopts a combination of exclusion methods having a high nighttime evaluation value depending on the kind of harmful animals at night when the near-infrared camera 21 is operated.

  The drone 2 carries out the removal method selected by the removal program.

<Third embodiment>
Next, a third embodiment will be described with reference to FIG. In addition, description is abbreviate | omitted about the structure which is common in 2nd embodiment, and it demonstrates centering on a different part from 2nd embodiment.

  The storage unit 52 of the drone 2 of the third embodiment has a capture success probability determination program. The capture success probability judgment program is a part of the judgment program. If the capture is not successful, the drone 2 calculates a success probability when the capture is performed again. The drone 2 operates based on the image acquired by the camera 20 or the camera 21, the operation of the removal target (movement speed, rate of change of the movement direction, etc.) and the performance of the capture net launching device 28 (follow-up performance to the movement of the removal target). And the success rate when the capture is performed again is calculated based on the shooting speed of the capture net 30 and the like. For example, when the movement speed of the removal target is faster than the follow-up performance of the capture net launching device 28, and the removal target is moving while greatly changing the movement direction, the success probability is low. On the other hand, if the operation to be disinfected is the same as or slower than the previous capture, and the possibility of capture is high if the firing conditions of the capture net 30 are corrected, the success probability is high. The drone 2 quantifies the success probability.

  If the success probability is equal to or higher than the predetermined value, the drone 2 performs the capture again by the removal program. The drone 2 calculates the launch conditions including the launch angle, launch direction, and launch speed of the capture net 30 based on the conditions when the success probability is calculated. On the other hand, if the success probability is less than a predetermined value, the drone 2 performs removal of the removal target. The predetermined value is, for example, 60%.

  Hereinafter, the operation of the drone 2 will be described with reference to FIG. When the drone 2 determines that the removal has not been completed in the case where the removal target is removed (step ST4 in FIG. 11), it determines whether the capture success probability is equal to or higher than a reference value (FIG. 16). Step ST41). If the capture success probability is equal to or higher than the reference value, the position of the extermination target is confirmed (step ST42), the correction value of the launch angle / direction and launch speed of the capture net 30 is calculated (step ST43), and the capture net 30 is fired. (Step ST44).

  On the other hand, if the capture success probability is not equal to or higher than the reference value, a combination of exclusion methods is selected (step ST45), and exclusion by the combination of exclusion methods is performed (step ST46).

  Note that the present invention is not limited to the above-described embodiments, and modifications, improvements, and the like within a scope in which the object of the present invention can be achieved are included in the present invention. For example, the base station 100 may perform part of the processing performed by the unmanned aircraft 1 (2) in the present embodiment. Moreover, you may combine the structure of the above-mentioned Embodiment 1, 2, and 3 suitably. Further, the evaluation value of the exclusion method combination may be updated by a predetermined method. Further, effective exclusion method combinations may be switched according to a predetermined standard. Alternatively, the drone 1 (2) can be dispatched as a single unit, a pattern in which a plurality of unmanned aircraft 1 (2) form a formation, or a pattern in which the drone 1 (2) is used in combination with an unmanned aerial vehicle is also excluded. It is good also as one of these. The unmanned aerial vehicle 1 (2) may be an unmanned air vehicle or an unmanned ship.

1, 2 drone 20 camera (visible light camera)
21 camera (near infrared camera)
24 light emitting device 26 sound generating device 28 capture net launching device 301 deer

Claims (10)

Autonomous movement means for autonomous movement;
Image acquisition means for acquiring an image in the intrusion prohibited area;
Determining means for recognizing whether an object in the image is a disinfection target, and determining whether the disinfection target has entered the intrusion prohibited area;
A disinfecting means for exterminating the disinfectant;
An unmanned aerial vehicle for pest control.   The pest control drone according to claim 1, wherein the extinguishing means includes a capturing means for capturing the extinguishing target and / or an excluding means for excluding the extinguishing target from the intrusion prohibited area. The exclusion method performed by the exclusion means includes sound generation, light irradiation, odor generation, movement in a specific movement mode, or a combination thereof.
An unmanned aerial vehicle for pest control according to claim 2. The exclusion method is defined for each type of the removal target,
The unmanned aircraft for pest control according to claim 3, wherein the unmanned aircraft for pest control activates the control means according to the type of the control target while tracking the control target. The exclusion method is defined by distinguishing between a daytime exclusion method and a nighttime exclusion method,
The pest control drone according to any one of claims 3 and 4, wherein the exclusion means is configured to selectively use the daytime exclusion method and the nighttime exclusion method.   The pest control drone according to any one of claims 1 to 5, wherein the image acquisition means is configured to acquire an image in visible light or an image in infrared according to the intensity of ambient light.   The pest according to any one of claims 1 to 6, wherein the determination unit refers to data generated by deep learning in order to recognize whether or not an object in the image is the target for extermination. A drone for disinfection. The determination unit determines whether the capture of the extermination target is successful based on the image acquired by the image acquisition unit. If it is determined that the capture is not successful, the determination unit performs the capture again. The probability of success
The said removal means performs the capture of the said removal object again if the success probability is more than a predetermined value, and implements the removal of the removal object if the success probability is less than the predetermined value. An unmanned aerial vehicle for pest control according to any one of 7 to 7. A pest control method implemented by an autonomously movable pest control drone,
An image acquisition step of acquiring an image in the intrusion prohibited area;
Recognizing whether or not the object in the image is a removal target, and determining whether or not the removal target has entered the intrusion prohibited area; and
A disinfecting step for disinfecting the disinfectant;
Pest control methods including: A computer that controls an autonomously movable pest control drone,
Image acquisition means for acquiring an image in the intrusion prohibited area;
Recognizing means for recognizing whether an object in the image is a disinfection target and determining whether the disinfection target has entered the intrusion prohibited area;
A disinfecting means for exterminating the disinfectant;
Program to function as.