Anomaly detection system, control device, anomaly detection method, and anomaly detection program

Abstract

The challenge is to efficiently achieve both comprehensive information gathering about the monitoring area and rapid and reliable detection of abnormal conditions. [Solution] The first drone 10 patrols the high altitude of the monitoring area (S1) and takes images of the monitoring area with its mounted camera. The second drone 20 patrols the low altitude of the monitoring area (S2) and takes images of the monitoring area with its mounted camera. If the first drone 10 detects an anomaly based on the images it has taken (S3), the second drone 20 circles at a low altitude near the location where the anomaly was detected to monitor the area (S4). If the second drone 20 determines that an anomaly has occurred based on the images it has taken, it notifies the monitoring terminal 40 of the anomaly information (S5).

Description

【Technical Field】 【0001】 The present invention relates to an abnormality detection system, a control device, an abnormality detection method, and an abnormality detection program that can efficiently achieve both rapid wide-range information collection and reliable detection of abnormal states for a wide monitoring area such as a power plant, a factory, or a theme park. 【Background Art】 【0002】 Conventionally, a technique of performing aerial photography using a flying object such as a drone and detecting an object from an image obtained thereby is known. For example, Patent Document 1 discloses a monitoring system that detects a shooting target such as a person or a vehicle from an image taken by a monitoring camera installed on the ground, moves a flying device according to the detected shooting target, and performs countermeasures such as warning an intruder and shooting a license plate. 【Prior Art Documents】 【Patent Documents】 【0003】 【Patent Document 1】 Japanese Patent Application Laid-Open No. 2016-118996 【Summary of the Invention】 【Problems to be Solved by the Invention】 【0004】 However, in the technology described in Patent Document 1 above, the target is detected from the image captured by the flying device. In such a system, it is conceivable to fly the flying device at a higher altitude to capture a wider area at once in order to collect information on the entire surveillance area in a short time. However, the higher the altitude at which the device flies, the smaller the target appears in the image, leading to problems such as detection failures where the target cannot be detected, and false detections where non-target objects are mistakenly identified as targets. To solve this problem, it is conceivable to fly at a lower altitude, but this narrows the shooting range, making it time-consuming to collect information on the entire surveillance area. Another method is to fly at a low altitude at high speed to collect information on a wider area in a short time, but when moving at high speed, there is a problem that the target may be missed, resulting in detection failures. 【0005】 The present invention was made to solve the problems (challenges) of the prior art described above, and aims to provide an anomaly detection system, control device, anomaly detection method, and anomaly detection program that can efficiently achieve both rapid and wide-area information gathering and reliable detection of abnormal conditions in a monitoring area. [Means for solving the problem] 【0006】 To solve the above problems, the present invention provides an anomaly detection system that includes a plurality of aircraft equipped with predetermined imaging devices and a control device for controlling the plurality of aircraft, and detects anomalies using images captured by the plurality of aircraft, wherein the control device includes a route designation unit that notifies a first aircraft of a first flight path and a second aircraft of a second flight path at a lower altitude than the first flight path, a first analysis unit that detects a target for detection based on a first image captured by the imaging device of the first aircraft flying along the first flight path, and a first determination unit that, if a target for detection is detected by the first analysis unit, gives a detailed monitoring instruction to at least the second aircraft for the location where the target for detection was detected. 【0007】 Furthermore, the present invention is characterized in that, in the above invention, the control device further comprises a second determination unit that determines whether or not there is an abnormality based on a second image captured by an imaging device of a second aircraft flying along the second flight path, and if it determines that there is no abnormality, it issues an instruction to the second aircraft to return to flight based on the second flight path. 【0008】 Furthermore, the present invention is characterized in that, in the above invention, the control device further comprises a second determination unit that determines whether or not there is an abnormality based on a second image captured by an imaging device of a second aircraft flying along the second flight path, and a notification unit that notifies abnormality information when the second determination unit determines that there is an abnormality. 【0009】 Furthermore, the present invention is a control device for controlling a plurality of aircraft equipped with a predetermined imaging device, comprising: a route designation unit that notifies a first aircraft of a first flight path and a second aircraft of a second flight path at a lower altitude than the first flight path; a first analysis unit that detects a target based on a first image captured by the imaging device of the first aircraft flying along the first flight path; and a first determination unit that, if a target is detected by the first analysis unit, gives at least the second aircraft a detailed monitoring instruction for the location where the target was detected. 【0010】 Furthermore, the present invention relates to an anomaly detection system comprising a plurality of aircraft equipped with predetermined imaging devices and a control device for controlling the plurality of aircraft, wherein an anomaly detection method is used to detect an anomaly using images captured by the plurality of aircraft, the method comprising: a route designation step in which the control device notifies a first aircraft of a first flight path and a second aircraft of a second flight path at a lower altitude than the first flight path; an analysis step in which a detection target is detected based on a first image captured by the imaging device of the first aircraft flying along the first flight path; and a determination step in which, if a detection target is detected by the analysis step, at least the second aircraft is instructed to perform detailed monitoring of the location where the detection target was detected. 【0011】 Furthermore, the present invention relates to an anomaly detection program executed by a control device that controls a plurality of aircraft equipped with a predetermined imaging device, characterized in that it causes a computer to execute a route designation procedure which notifies a first aircraft of a first flight path and a second aircraft of a second flight path at a lower altitude than the first flight path; an analysis procedure which detects a target for detection based on a first image captured by the imaging device of the first aircraft flying along the first flight path; and a determination procedure which, if a target for detection is detected by the analysis procedure, at least the second aircraft is instructed to perform detailed monitoring of the location where the target for detection was detected. [Effects of the Invention] 【0012】 According to the present invention, it is possible to efficiently achieve both rapid and wide-ranging information gathering about the monitoring area and reliable detection of abnormal conditions. [Brief explanation of the drawing] 【0013】 [Figure 1] Figure 1 is an explanatory diagram illustrating the overview of the anomaly detection system according to the embodiment. [Figure 2] Figure 2 shows the system configuration of the anomaly detection system according to the embodiment. [Figure 3] Figure 3 is a functional block diagram showing the configuration of the first drone shown in Figure 2. [Figure 4] Figure 4 shows an example of the flight path data shown in Figure 3. [Figure 5] Figure 5 is a functional block diagram showing the configuration of the second drone shown in Figure 2. [Figure 6] Figure 6 shows an example of the flight path data shown in Figure 5. [Figure 7] Figure 7 is a functional block diagram showing the configuration of the control device shown in Figure 2. [Figure 8] Figure 8 shows an example of the first flight path data, second flight path data, and anomaly location data shown in Figure 7. [Figure 9]FIG. 9 is a flowchart showing a processing procedure related to drone control according to an embodiment. [Figure 10] FIG. 10 is an explanatory diagram of the outline of the abnormality detection system according to Modification 1. [Figure 11] FIG. 11 is an explanatory diagram (part 1) of the outline of the abnormality detection system according to Modification 2. [Figure 12] FIG. 12 is an explanatory diagram (part 2) of the outline of the abnormality detection system according to Modification 2. [Figure 13] FIG. 13 is an explanatory diagram (part 3) of the outline of the abnormality detection system according to Modification 2. [Figure 14] FIG. 14 is an explanatory diagram of the outline of the abnormality detection system according to Modification 3. [Figure 15] FIG. 15 is an explanatory diagram of the outline of the abnormality detection system according to Modification 4. [Figure 16] FIG. 16 is an explanatory diagram (part 1) of the outline of the abnormality detection system according to Modification 5. [Figure 17] FIG. 17 is an explanatory diagram (part 2) of the outline of the abnormality detection system according to Modification 5. [Figure 18] FIG. 18 is an explanatory diagram (part 3) of the outline of the abnormality detection system according to Modification 5. Embodiments for Carrying Out the Invention 【0014】 <s [Embodiment] Hereinafter, embodiments of the abnormality detection system, control device, abnormality detection method, and abnormality detection program according to the present embodiment will be described in detail based on the drawings. 【0015】 <Outline of the Abnormality Detection System According to the Embodiment> First, the outline of the abnormality detection system according to the present embodiment will be described. FIG. 1 is an explanatory diagram of the outline of the abnormality detection system according to the present embodiment. In the abnormality detection system according to the present embodiment, two drones (corresponding to the "flying object" described in the claims) are operated to monitor a monitoring area. 【0016】 As shown in Figure 1(a), the first drone 10 flies (patrols) at a high altitude above the monitoring area (hereinafter referred to as "high altitude") according to a pre-set flight path (S1), and captures images of the monitoring area with its mounted camera. The second drone 20 flies (patrols) at a lower altitude above the monitoring area than the first drone 10 (hereinafter referred to as "low altitude") according to a pre-set flight path (S2), and captures images of the monitoring area with its mounted camera (hereinafter, the act of flying and capturing images may be collectively referred to as "patrol monitoring"). 【0017】 As shown in Figure 1(b), if the first drone 10 detects an intruding object such as a person or vehicle, or flames or smoke (hereinafter collectively referred to as "detection target") in the image it has captured (S3), the second drone 20 circles at a low altitude near the point where the first drone 10 captured the image of the detection target and captures the monitoring area (S4). 【0018】 If the image captured by the second drone 20 detects an object and determines that an anomaly has occurred, the anomaly information is sent to the monitoring terminal 40 (S5). 【0019】 As described above, in the anomaly detection system according to this embodiment, if a first drone that patrols and monitors at high altitude detects a target from the image it has captured and determines that there is an anomaly, a second drone circles and monitors the area above the point where the first drone detected the target at low altitude. If the second drone captures an image in which it detects a target and determines that there is an anomaly, the system is configured to notify the monitoring operator's terminal of the anomaly information. 【0020】 As a result, even if the images captured by the first drone from high altitude do not provide sufficient image information to determine an anomaly, and something that is not the target is detected as an anomaly, the second drone captures images from low altitude at the location where the first drone detected the target, obtaining more detailed image information. If it is not the target, i.e., a false detection, the system does not notify the monitoring terminal of the anomaly information, thus enabling reliable detection of abnormal conditions. Furthermore, the first drone patrols the airspace above the monitoring area, capturing images over a wide area from high altitude, while the second drone, patrolling at a lower altitude, rotates and monitors as needed to capture detailed images. This allows for efficient and rapid collection of information over a wide area of ​​the monitoring region, as well as reliable detection of abnormal conditions. 【0021】 <System configuration of the anomaly detection system according to the embodiment> Next, the system configuration of the anomaly detection system according to this embodiment will be described. Figure 2 is a diagram showing the system configuration of the anomaly detection system according to this embodiment. 【0022】 As shown in Figure 2, the first drone 10 and the second drone 20 are connected to the network N via wireless communication, and are communicated with the management device 30 via this network N. 【0023】 Furthermore, the monitoring terminal 40 connects to network N via wireless communication, and is connected to the management device 30 via network N in a communicative manner. Network N may be a network dedicated to the anomaly detection system, or it may be a public network such as the Internet. 【0024】 The first drone 10 and the second drone 20 are drones for monitoring the monitoring area and capture images of the monitoring area with their mounted cameras. When the first drone 10 and the second drone 20 receive flight path data from the management device 30, they store the received flight path data. 【0025】 Furthermore, when the first drone 10 and the second drone 20 receive a patrol commencement instruction from the management device 30, they perform a flight based on the stored flight path data. 【0026】 Furthermore, the first drone 10 and the second drone 20 calculate their position information using positioning signals received from satellites in a GNSS (Global Navigation Satellite System) such as GPS. They then associate this position information with image information captured by their onboard cameras and notify the management device 30. 【0027】 Furthermore, if the first drone 10 receives a detailed monitoring instruction from the management device 30, it will hover above the location information included in the detailed monitoring instruction. If the second drone 20 receives a detailed monitoring instruction from the management device 30, it will orbit around the location information included in the detailed monitoring instruction. 【0028】 Furthermore, if the first drone 10 and the second drone 20 receive a patrol return instruction from the management device 30, they will resume patrol flight based on the flight path data. If the first drone 10 and the second drone 20 receive a patrol end instruction from the management device 30, they will land at a predetermined location. 【0029】 The management device 30 is a device that performs flight control and abnormality detection processing for the first drone 10 and the second drone 20. When the management device 30 receives route information for the patrol flight of the first drone 10 and the second drone 20 in response to the operator's input, it notifies the first drone 10 and the second drone 20, respectively, of the received route information as flight path data. 【0030】 Furthermore, if the control device 30 receives a patrol commencement instruction in response to an operator's input, it notifies the first drone 10 and the second drone 20 of this patrol commencement instruction. 【0031】 Furthermore, when the management device 30 receives image information associated with location information (hereinafter referred to as "first image information") from the first drone 10, it processes the image information contained in the received first image information to extract detection targets such as people, vehicles, and smoke, and also extracts the location information associated with this image and stores it in the anomaly location data. Then, it determines whether or not there is an anomaly regarding the presence of the extracted people or vehicles. If an anomaly is determined in this determination, it notifies the first drone 10 and the second drone 20 of a detailed monitoring instruction including the anomaly location data. 【0032】 Furthermore, when the management device 30 receives image information associated with location information (hereinafter referred to as "second image information") from the second drone 20, it processes the image information contained in the received second image information to extract the image to be detected, and also extracts the location information associated with this image, stores it in the anomaly location data, and updates it. Then, it determines whether or not there is an anomaly regarding the presence of the extracted person or vehicle. If an anomaly is determined in this determination, it notifies the monitoring terminal 40 of the anomaly information, including the corresponding image information and anomaly location data. 【0033】 Furthermore, after notifying the operator of detailed monitoring instructions, if the management device 30 does not extract an image of the target to be detected from the second image information, or if an image of the target to be detected is extracted but not determined to be abnormal, while the second drone 20 is circling, it notifies the first drone 10 and the second drone 20 of a return to patrol instructions. 【0034】 Furthermore, if the management device 30 receives a patrol return instruction or a patrol end instruction in response to the operator's actions, it notifies the first drone 10 and the second drone 20 of the received instruction. 【0035】 The monitoring terminal 40 is a terminal carried by the monitoring officer. When the monitoring terminal 40 receives abnormal information from the management device 30, it displays this abnormal information. When displaying abnormal information, it may sound an alarm or illuminate an LED or other light source. 【0036】 <Configuration of the first drone 10> Next, the configuration of the first drone 10 shown in Figure 2 will be described. Figure 3 is a functional block diagram showing the configuration of the first drone 10 shown in Figure 2. As shown in Figure 3, the first drone 10 has a GNSS receiver 15, a camera 16, a motor 17, and a control device 11. 【0037】 The GNSS receiver 15 is a receiving device that receives positioning signals from satellites in GNSS such as GPS. The camera 16 is an imaging device for imaging the monitoring area. The motor 17 is a drive device for rotating the propeller. 【0038】 The control device 11 is a device that controls the first drone 10 and includes a control unit 12, a storage unit 13, and a communication unit 14. The communication unit 14 is an interface unit for data communication with the network N using wireless communication. 【0039】 The memory unit 13 is a storage device such as a hard disk drive or non-volatile memory, and stores the flight path data 13a. The flight path data 13a is data indicating the route that the first drone 10 will fly, and includes coordinates and altitude. The coordinates are expressed in latitude and longitude. 【0040】 The control unit 12 is a control unit that performs overall control of the control device 11 and includes a flight path acquisition unit 12a, a flight control unit 12b, a position information acquisition unit 12c, an image information acquisition unit 12d, and an information notification unit 12e. In practice, these programs are loaded into the CPU (Central Processing Unit) and executed, causing the processes corresponding to the flight path acquisition unit 12a, the flight control unit 12b, the position information acquisition unit 12c, the image information acquisition unit 12d, and the information notification unit 12e to execute, respectively. 【0041】 The flight path acquisition unit 12a is a processing unit that manages the flight path data 13a. When the flight path acquisition unit 12a receives flight path data from the management device 30, it stores the received data in the flight path data 13a of the storage unit 13. 【0042】 The flight control unit 12b is a processing unit that controls the flight of the first drone 10. When the flight control unit 12b receives a patrol start instruction from the management device 30, it controls the motor 17 to perform a flight based on the flight path data 13a. 【0043】 Furthermore, if the flight control unit 12b receives a detailed monitoring instruction from the management device 30, it controls the motor 17 to hover above the location information included in the detailed monitoring instruction. 【0044】 Furthermore, if the flight control unit 12b receives a command to return to patrol flight from the management device 30, it controls the motor 17 to return to patrol flight based on the flight path data 13a. 【0045】 Furthermore, if the flight control unit 12b receives a patrol completion instruction from the management device 30, it controls the motor 17 to land at the predetermined position. 【0046】 The location information acquisition unit 12c is a processing unit that acquires the location information of the first drone 10. The location information acquisition unit 12c calculates the location information using the positioning signal received by the GNSS receiver unit 15. 【0047】 The image information acquisition unit 12d is a processing unit that acquires image information of the monitoring area. The image information acquisition unit 12d acquires the image captured by the camera 16 as image information. 【0048】 The information notification unit 12e is a processing unit that notifies location information and image information. When location information is calculated by the location information acquisition unit 12c, the information notification unit 12e associates this location information with the image information acquired by the image information acquisition unit 12d and notifies the management device 30. 【0049】 Next, an example of data stored in the memory unit 13 of the control device 11 shown in Figure 3 will be described. Figure 4 shows an example of the flight path data 13a shown in Figure 3. 【0050】 The flight path data 13a shown in Figure 4 associates the coordinates "35.xx046, 139.xx240" with an altitude of "30" m, and the coordinates "35.xx035, 139.xx262" with an altitude of "30" m. 【0051】 <Configuration of the second drone 20> Next, the configuration of the second drone 20 shown in Figure 2 will be described. Figure 5 is a functional block diagram showing the configuration of the second drone 20 shown in Figure 2. As shown in Figure 5, the second drone 20 has a GNSS receiver 15, a camera 16, a motor 17, and a control device 21. Note that the explanation of the functional parts, which are the same as those of the first drone 10 shown in Figure 3, will be omitted. 【0052】 The control device 21 is a device that controls the second drone 20 and includes a control unit 22, a storage unit 23, and a communication unit 14. 【0053】 The memory unit 23 is a storage device such as a hard disk drive or non-volatile memory, and stores the flight path data 23a. The flight path data 23a is data indicating the route that the second drone 20 will fly. 【0054】 The control unit 22 is a control unit that performs overall control of the control device 21 and includes a flight path acquisition unit 22a, a flight control unit 22b, a position information acquisition unit 12c, an image information acquisition unit 12d, and an information notification unit 12e. In practice, by loading these programs into the CPU and executing them, the processes corresponding to the flight path acquisition unit 22a, the flight control unit 22b, the position information acquisition unit 12c, the image information acquisition unit 12d, and the information notification unit 12e are made to execute, respectively. 【0055】 The flight path acquisition unit 22a is a processing unit that manages the flight path data 23a. When the flight path acquisition unit 22a receives flight path data from the management device 30, it stores the received data in the flight path data 23a of the storage unit 23. 【0056】 The flight control unit 22b is a processing unit that controls the flight of the second drone 20. When the flight control unit 22b receives a patrol start instruction from the management device 30, it controls the motor 17 to perform a flight based on the flight path data 23a. 【0057】 Furthermore, if the flight control unit 22b receives a detailed monitoring instruction from the management device 30, it controls the motor 17 to orbit around the position specified by the position information included in the detailed monitoring instruction. 【0058】 Furthermore, if the flight control unit 22b receives a return-to-patrol instruction from the management device 30, it controls the motor 17 to return to patrol flight based on the flight path data 23a. 【0059】 Furthermore, if the flight control unit 22b receives a patrol completion instruction from the management device 30, it controls the motor 17 to land at the predetermined position. 【0060】 Next, an example of data stored in the memory unit 23 of the control device 21 shown in Figure 5 will be described. Figure 6 shows an example of the flight path data 23a shown in Figure 5. 【0061】 The flight path data 23a shown in Figure 6 associates the coordinates "35.xx046, 139.xx240" with an altitude of "10" m, and the coordinates "35.xx035, 139.xx262" with an altitude of "10" m. 【0062】 <Configuration of the control device 30> Next, the configuration of the management device 30 shown in Figure 2 will be described. Figure 7 is a functional block diagram showing the configuration of the management device 30 shown in Figure 2. As shown in Figure 7, the management device 30 is connected to the display unit 35 and the input unit 36 ​​and has a control unit 32, a storage unit 33, and a communication unit 34. 【0063】 The display unit 35 is a display device such as an LCD panel or a display device. The input unit 36 ​​is an input device such as a keyboard or mouse. The communication unit 34 is an interface unit for data communication between the first drone 10, the second drone 20 and the monitoring terminal 40 via the network N. 【0064】 The memory unit 33 is a storage device such as a hard disk drive or non-volatile memory, and stores first flight path data 33a, second flight path data 33b, first image data 33c, second image data 33d, and anomaly location data 33e. 【0065】 The first flight path data 33a is data showing the route taken by the first drone 10 during its patrol flight. The second flight path data 33b is data showing the route taken by the second drone 20 during its patrol flight. 【0066】 The first image data 33c is data showing image information captured by the first drone 10. The second image data 33d is data showing image information captured by the second drone 20. The anomaly location data 33e is data showing the coordinates of the location where the detected object was found. 【0067】 The control unit 32 is a control unit that performs overall control of the management device 30, and includes a first flight path designation unit 32a, a second flight path designation unit 32b, a drone control unit 32c, a first image analysis unit 32d, a first position information acquisition unit 32e, a first determination unit 32f, a second image analysis unit 32g, a second position information acquisition unit 32h, a second determination unit 32i, and an anomaly notification unit 32j. In practice, by loading these programs into the CPU and executing them, the processes corresponding to the first flight path designation unit 32a, the second flight path designation unit 32b, the drone control unit 32c, the first image analysis unit 32d, the first position information acquisition unit 32e, the first determination unit 32f, the second image analysis unit 32g, the second position information acquisition unit 32h, the second determination unit 32i, and the anomaly notification unit 32j are executed, respectively. 【0068】 The first flight path designation unit 32a is a processing unit that manages and notifies the first flight path data 33a. When the first flight path designation unit 32a receives route information for the first drone 10's patrol flight from the input unit 36, it stores the received route information in the first flight path data 33a and notifies the first drone 10 of the flight path data. 【0069】 The second flight path designation unit 32b is a processing unit that manages and notifies the second flight path data 33b. When the second flight path designation unit 32b receives route information for the second drone 20's patrol flight from the input unit 36, it stores the received route information in the second flight path data 33b and notifies the second drone 20 of the flight path data. 【0070】 The drone control unit 32c is a processing unit that controls the first drone 10 and the second drone 20. When the drone control unit 32c receives a patrol start instruction from the input unit 36 ​​in response to the operator's operation, it notifies the first drone 10 and the second drone 20 of this patrol start instruction. In this embodiment, a configuration in which a patrol start instruction is notified in response to the operator's operation is described, but the system is not limited to this, and the patrol start instruction may be notified according to a pre-set schedule. In addition, when an intruder is detected by the security system, the system may set patrol flight route information according to the detection status of the sensors constituting the security system and notify it as flight path data, and also notify the patrol start instruction. 【0071】 Furthermore, if the first determination unit 32f determines that an abnormality has occurred, the drone control unit 32c notifies the first drone 10 and the second drone 20 of a detailed monitoring instruction including abnormality location data 33e. 【0072】 Furthermore, if the second determination unit 32i determines that the drone control unit 32c is functioning normally, it notifies the first drone 10 and the second drone 20 of a patrol return instruction. 【0073】 Furthermore, if the drone control unit 32c receives a patrol return instruction from the input unit 36, it notifies the first drone 10 and the second drone 20 of this patrol return instruction. 【0074】 Furthermore, if the drone control unit 32c receives a patrol completion instruction from the input unit 36, it notifies the first drone 10 and the second drone 20 of this patrol completion instruction. 【0075】 The first image analysis unit 32d is a processing unit that analyzes image information received from the first drone 10. When the first image analysis unit 32d receives image information associated with location information from the first drone 10, it stores the received image information in the first image data 33c. Then, it processes the received image information to extract images of detection targets such as people and vehicles. 【0076】 The first position information acquisition unit 32e is a processing unit that acquires position information from the first drone 10. If the first image analysis unit 32d extracts an image of the target to be detected, the first position information acquisition unit 32e extracts the position information associated with the extracted image information of the target to be detected from the first image data 33c and stores it in the abnormal location data 33e. 【0077】 The first determination unit 32f is a processing unit that determines whether or not there is an abnormality based on the analysis results of the image information stored in the first image data 33c. If the first image analysis unit 32d extracts an image of the object to be detected, the first determination unit 32f determines whether or not there is an abnormality regarding the presence of the extracted object at the location indicated in the abnormal location data 33e. For example, if the first image analysis unit 32d detects a parked vehicle and the location indicated in the abnormal location data 33e is not a parking lot, the first determination unit 32f determines that there is an abnormality because the vehicle parked at this location may be a suspicious vehicle. In other words, the first determination unit 32f determines that there is an abnormality when an object to be detected, such as a person or vehicle, is detected in a place where it is not normally detected. 【0078】 The second image analysis unit 32g is a processing unit that analyzes image information received from the second drone 20. When the second image analysis unit 32g receives image information associated with location information from the second drone 20, it stores the received image information in the second image data 33d. Then, it processes the received image information to extract images of detection targets such as people and vehicles. 【0079】 The second location information acquisition unit 32h is a processing unit that acquires location information from the second drone 20. If the second image analysis unit 32g extracts an image of the target to be detected, the second location information acquisition unit 32h extracts the location information associated with the extracted image information of the target to be detected from the second image data 33d and stores it in the abnormal location data 33e. 【0080】 The second determination unit 32i is a processing unit that determines whether or not there is an abnormality based on the analysis results of the image information stored in the second image data 33d. If the second image analysis unit 32g extracts an image of the target to be detected, the second determination unit 32i determines whether or not there is an abnormality regarding the presence of the extracted target at the location indicated in the abnormal location data 33e. For example, if the second image analysis unit 32g detects a parked vehicle, it compares the vehicle's appearance and license plate information with information on vehicles pre-registered as vehicles of facility users to determine whether or not the vehicle is permitted to be parked as a vehicle of a facility user, and whether or not the location indicated in the abnormal location data 33e is a designated parking lot where parking is permitted in advance. In addition, if the second image analysis unit 32g detects a person, it performs authentication to determine if the person is a facility user based on the person's facial image and what the person is wearing (uniform, employee ID, etc.). If the authentication result is that the person is not a facility user, and the location indicated in the abnormal location data 33e is not a space where people other than facility users are permitted to enter and exit, it determines that there is an abnormality as the person is a suspicious person. 【0081】 Furthermore, the second determination unit 32i determines that the situation is normal if, after the drone control unit 32c notifies the second drone 20 of a detailed monitoring instruction and before the second drone 20 completes its circling flight, the second image analysis unit 32g does not extract an image of the target to be detected, or if the extracted target to be detected is not determined to be abnormal. The second determination unit 32i also determines whether the second drone 20 has completed its circling flight based on the position information of the second image data 33d. 【0082】 The abnormality notification unit 32j is a processing unit that notifies abnormality information. If the second determination unit 32i determines that there is an abnormality, the abnormality notification unit 32j notifies the monitoring terminal 40 of the abnormality information, including the image information and abnormality location data 33e used in the determination. 【0083】 Next, an example of data stored in the storage unit 33 of the management device 30 shown in Figure 7 will be described. Figure 8 shows an example of the first flight path data 33a, second flight path data 33b, and anomaly location data 33e shown in Figure 7. 【0084】 The first flight path data 33a shown in Figure 8(a) associates the coordinates "35.xx046, 139.xx240" with an altitude of "30" m, and the coordinates "35.xx035, 139.xx262" with an altitude of "30" m. 【0085】 The second flight path data 33b shown in Figure 8(b) associates the coordinates "35.xx046, 139.xx240" with an altitude of "10" m, and the coordinates "35.xx035, 139.xx262" with an altitude of "10" m. 【0086】 The anomaly location data 33e shown in Figure 8(c) indicates a state where the coordinates are "35.xx040, 139.xx252". 【0087】 <Processing procedure related to drone control according to the embodiment> Next, the processing procedure for drone control according to this embodiment will be described. Figure 9 is a flowchart showing the processing procedure for drone control according to this embodiment. 【0088】 When the management device 30 receives route information for the patrol flight of the first drone 10 via an operator's command, it stores the received route information as flight path data and notifies the first drone 10 of the flight path data. When the first drone 10 receives the flight path data, it stores this flight path data. Similarly, when the management device 30 receives route information for the patrol flight of the second drone 20 via an operator's command, it stores the received route information as flight path data and notifies the second drone of the flight path data. When the second drone 20 receives the flight path data, it stores this flight path data. If the first drone 10 and the second drone 20 fly the same flight path at different altitudes, the management device 30 may, upon receiving the coordinates (latitude, longitude) indicating the flight path via an operator's command, store the flight paths with the same coordinates (latitude, longitude) but different altitudes as route information and notify both the first drone 10 and the second drone 20. As shown in Figure 9, when the control device 30 receives a patrol start instruction from the operator, it notifies the first drone 10 and the second drone 20 of the patrol start instruction. If the management device 30 notifies the first drone 10 and the second drone 20 of the instruction to start patrolling (step S101), the first drone 10 starts patrolling based on the stored flight path data, captures an image of the monitoring area, and notifies the management device 30 of the first image information associated with the location information (step S102). The second drone 20 also starts patrolling based on the stored flight path data, captures an image of the monitoring area, and notifies the management device 30 of the second image information associated with the location information (step S103). 【0089】 If the management device 30 does not detect any abnormalities in the received first image information (step S104; No), it continues the detection process in step S104 until the patrol monitoring by the first drone 10 and the second drone 20 is completed or a target for detection is detected. If an abnormality is detected in the received first image information (step S104; Yes), it notifies the first drone 10 and the second drone 20 of a detailed monitoring instruction (step S105). 【0090】 When the first drone 10 receives a detailed monitoring instruction, it begins to hover above the location identified by the anomaly location data included in the detailed monitoring information, that is, the location where the detection target was detected (step S106). When the second drone 20 receives a detailed monitoring instruction, it begins to circle around the location where the detection target identified by the anomaly location data included in the detailed monitoring information was detected (step S107), and notifies the management device 30 of the second image information associated with the location information (step S108). 【0091】 If the management device 30 determines that it has not detected the target object in the received second image information, or has detected the target object but there is no abnormality (step S109; No), and the rotation by the second drone has been completed (step S110; Yes), it proceeds to step S113. If the rotation by the second drone has not been completed (step S110; No), it returns to step S109. 【0092】 If the management device 30 detects the target in the received second image information and determines that there is an abnormality (step S109; Yes), it notifies the monitoring terminal 40 of the abnormality (step S111). If it receives a patrol return instruction from the operator (step S112; Yes), it proceeds to step S113. If it has not received a patrol return instruction (step S112; No), it returns to step S112. 【0093】 When the management device 30 receives a patrol return instruction (step S112; Yes), it sends a patrol return instruction to the first drone 10 and the second drone 20 respectively (step S113), returns to step S104, and performs detection processing of the target to be detected until the patrol monitoring by the first drone 10 and the second drone 20 is completed. Upon receiving the patrol return instruction from the management device 30, the first drone 10 and the second drone return to patrol monitoring (steps S114, S115) and terminate processing. 【0094】 As described above, in the anomaly detection system according to this embodiment, if the first drone patrolling at high altitude determines that there is an anomaly from the image it has captured, the second drone patrolling at low altitude will circle and monitor the area above the location where the anomaly was determined to be present. If the second drone determines that there is an anomaly from the image it has captured, it will notify the monitoring operator's terminal of the anomaly information. This configuration enables rapid and wide-area information gathering of the monitoring area using the image captured by the first drone, while efficiently achieving reliable detection of anomalies using more detailed images captured by the second drone. 【0095】 In the above embodiment, a configuration was described in which, if the management device determines that there is an abnormality using the image information captured by the first drone, it instructs the second drone to orbit and monitor the location where the abnormality was detected. However, the present invention is not limited to this. The first drone can also be configured to determine an abnormality using the image information captured by its own device, and if an abnormality is detected, the first drone can directly instruct the second drone to orbit and monitor the location where the abnormality was detected. 【0096】 Furthermore, although the above embodiment described a configuration in which two drones are used to monitor the surveillance area, the present invention is not limited thereto, and it is also possible to configure the system to use three or more drones to monitor the surveillance area. In this case, in addition to high-altitude and low-altitude flights, drones may be operated individually for roles such as flying around important locations such as research facilities, or each drone may be assigned a role, including standby, on a rotational basis. In the above embodiment, it was explained that the first drone and the second drone are equipped with cameras to capture images, but the cameras are not limited to visible light cameras; any camera capable of detecting an object from the acquired information, such as an infrared camera or a distance sensor, is acceptable. 【0097】 Furthermore, in the above embodiment, a configuration was described in which, when the management device determines that there is an abnormality using the image information captured by the first drone, it instructs the second drone to orbit and monitor the location where the abnormality was determined. However, the present invention is not limited to this. When the management device determines that there is an abnormality using the image information captured by the first drone, it can also be configured to predict the destination of the object to be detected based on its movement and instruct the second drone to orbit and monitor the area above the predicted destination. 【0098】 [Example 1] By the way, in the above embodiment, when the first drone, which patrols and monitors at high altitude, determines that there is an anomaly based on the image it has captured, the first drone hovers at high altitude while the second drone circles and monitors the location where the anomaly was determined to be present at low altitude. However, the present invention is not limited to this. 【0099】 If the first drone, which is patrolling and monitoring at high altitude, detects an anomaly based on the images it has captured, the second drone may be instructed to circle and monitor the location where the anomaly was detected at low altitude, while the first drone continues its patrol and monitoring at high altitude. Furthermore, if the first drone detects a new anomaly based on the images it has captured, the second drone may be configured to move to the location where the anomaly was detected via the shortest route. 【0100】 This modified example 1 describes an anomaly detection system in which, if a first drone patrolling and monitoring at high altitude determines that an anomaly is present based on images it has captured, a second drone is instructed to circle and monitor the location where the anomaly was detected at low altitude, while the first drone continues its patrol and monitoring at high altitude. If the first drone detects a new anomaly from images it has captured, the second drone can move to the location where the anomaly was detected by taking the shortest possible route. 【0101】 <Overview of the anomaly detection system related to the modified example 1> The outline of the anomaly detection system according to this modified example 1 will be explained. Figure 10 is an explanatory diagram of the outline of the anomaly detection system according to this modified example 1. 【0102】 As shown in Figure 10(a), in this anomaly detection system, if the first drone 110, which is patrolling at high altitude, determines that there is an anomaly based on the images it has captured (S11), the management device 30 notifies the second drone 120, which is patrolling at low altitude, of a detailed monitoring instruction. Upon receiving the detailed monitoring instruction, the second drone 20 circles at low altitude and monitors the area above the point identified by the anomaly location data included in the detailed monitoring instruction (S12). The first drone 110 continues to patrol and monitor along a predetermined flight path based on the flight path data (S13). 【0103】 Subsequently, as shown in Figure 10(b), if the first drone 110, which is patrolling and monitoring at high altitude, determines that there is a new anomaly based on the images it has captured (S14), the second drone 120 moves along the shortest route to the vicinity of the location where the new anomaly was determined. At this time, if there are obstacles such as buildings in the path of movement, the second drone 120 flies to avoid these obstacles (S15). 【0104】 As described above, in the anomaly detection system according to this modified example 1, if the first drone, which patrols and monitors at high altitude, determines that there is an anomaly based on the images it has captured, the second drone will circle and monitor the location where the anomaly was determined at low altitude, while the first drone will continue its patrol and monitoring at high altitude. If the first drone captures images and determines that there is a new anomaly, the second drone will move to the location where the anomaly was newly determined to be present, taking the shortest distance. This configuration allows for efficient and rapid collection of information over a wide area of ​​the monitoring region, as well as reliable detection of anomalies. 【0105】 [Differentiation 2] By the way, in the above embodiment, when a first drone that patrols and monitors at high altitude determines that there is an anomaly based on the image it has captured, the first drone hovers at high altitude while the second drone circles and monitors the location where the anomaly was determined to be present at low altitude. However, the present invention is not limited to this. 【0106】 The system can also be configured to determine that there is an "abnormality" if the person or object detected by the first drone from the image captured by the first drone (hereinafter referred to as the "detected object") is of the same type as the detected object detected by the second drone from the image captured by the second drone, or if the second drone detects the same type of detected object a predetermined number of times from the image captured by the second drone. Furthermore, if the second drone detects an object that is different from the detected object detected by the first drone from the image captured by the first drone, but has a similar shape, the system can be configured to determine that the detection by the first drone was a "false detection". 【0107】 This modified example 2 describes an anomaly detection system in which, if the object detected from the image captured by the first drone is the same type as the object detected from the image captured by the second drone, or if the object detected from the image captured by the second drone is the same type and is detected a predetermined number of times, a final "anomaly" is determined. Alternatively, if the object detected from the image captured by the first drone is different from the object detected from the image captured by the first drone but has a similar shape, the detection from the image captured by the second drone is determined to be a "false detection." 【0108】 <Overview of the anomaly detection system related to Modification Example 2> The outline of the anomaly detection system according to this modified example 2 will be explained. Figures 11 to 13 are explanatory diagrams illustrating the outline of the anomaly detection system according to this modified example 2. 【0109】 In this anomaly detection system, if the object detected from the image captured by the first drone and the object detected from the image captured by the second drone are of the same type, it is ultimately determined that there is an anomaly. 【0110】 For example, Figure 11(a) shows an image in which an anomaly is detected based on images taken by the first drone patrolling and monitoring at high altitude. In Figure 11(a), a vehicle is detected as the target object. 【0111】 Figure 11(b) shows an image captured by a second drone that circled at low altitude near the location where an anomaly was detected. As shown in Figure 11(b), the second drone detected the same type of vehicle as the vehicle detected in the image captured by the first drone. In this case, the final determination is "anomaly detected". 【0112】 Figure 11(c) shows the image captured by the second drone, which reveals tire tracks left by a vehicle. These tracks are determined to be traces left by the vehicle that was captured in the image taken by the first drone. In this case, although the vehicle is not visible in the image taken by the second drone, tire tracks indicating the presence of a vehicle were detected. Therefore, it is determined that a similar vehicle-like object was detected, and the final result is "anomaly detected." 【0113】 Furthermore, if the object detected by the first drone is a person, and the second drone can detect traces of the person's passage, such as a broken fence, torn barbed wire, or destroyed wall, from the images it captures, then it may be determined that a similar object indicating a person has been detected, and ultimately judged as "abnormal." 【0114】 Furthermore, in this anomaly detection system, if an object detected from an image captured by the second drone is the same type as an object detected from an image captured by the first drone, and this is detected a predetermined number of times (for example, twice), it is ultimately determined to be an "anomaly." 【0115】 For example, if the object detected from the image captured by the first drone is identified as a vehicle, the images captured by the second drone are shown in Figures 12(a) to 12(c). In the images in Figures 12(a) and 12(c), the presence of the vehicle can be recognized, but in the image in Figure 12(b), it is difficult to recognize the presence of the vehicle. In Figures 12(a) to 12(c), the vehicle is detected twice, and because it was detected twice, the predetermined number of times, by the second drone's rotation monitoring, it is ultimately determined that there is an "abnormality". 【0116】 Furthermore, in this anomaly detection system, if the second drone detects an object with a similar shape to the object detected by the first drone using images, even if it is different from the object detected by the first drone using images, the detection by the first drone will be considered a false positive and the system will determine that there is "no anomaly." 【0117】 For example, Figure 13(a) shows an image in which an anomaly is detected based on an image taken by the first drone patrolling and monitoring at high altitude. In Figure 13(a), the image is determined to be a person taken from directly above, and the person is detected as the object to be detected. 【0118】 Figure 13(b) shows an image captured by a second drone that circled at low altitude near the location where an anomaly was detected. As shown in Figure 13(b), this image shows a fire hydrant. When a fire hydrant is photographed from above, the resulting image is similar to that of a person photographed from above, as shown in Figure 13(a). Therefore, it is thought that the first drone mistakenly detected the fire hydrant as a person in the image it captured. In other words, the detection by the first drone is considered a false positive, and the final determination is "no anomaly." 【0119】 As described above, in the anomaly detection system according to this modified example 2, if the object detected from the image captured by the first drone and the object detected from the image captured by the second drone are of the same type, or if the object detected from the image captured by the second drone is of the same type and is detected a predetermined number of times, it is ultimately determined that there is an "anomaly." Furthermore, if the object detected from the image captured by the second drone is different from the object detected from the image captured by the first drone but has a similar shape, the detection by the first drone is considered a false positive and it is determined that there is "no anomaly." With this configuration, it is possible to efficiently achieve both rapid and wide-ranging information gathering about the monitoring area and reliable detection of anomaly conditions. 【0120】 In the above modified example 2, a configuration was described in which a final "abnormality detected" is determined when the detected object from the image captured by the first drone and the detected object from the image captured by the second drone are of the same type. However, the present invention is not limited to this. When it is difficult to determine conditions such as when the detected object from the image captured by the first drone and the detected object from the image captured by the second drone are of the same type, the system can also be configured to notify the monitoring terminal of a "possible abnormality." 【0121】 [Difference 3] By the way, in the above embodiment, when a first drone patrolling and monitoring at high altitude determines that there is an anomaly based on the image it has captured, a second drone patrolling and monitoring at low altitude circles and monitors the area above the location where the anomaly was determined to be present, and if the second drone determines that there is an anomaly based on the image it has captured, it notifies the monitoring officer's terminal of the anomaly information. However, the present invention is not limited to this. 【0122】 If the object detected by the first drone based on the image it has captured is determined to be in a clearly abnormal state, the system can be configured to notify the monitoring terminal of the abnormality information without requiring further judgment using the image captured by the second drone. 【0123】 This modified example 3 describes an anomaly detection system that, if an object detected based on an image taken by the first drone is determined to be in a clearly abnormal state, notifies the monitoring terminal of the anomaly information without performing a determination using an image taken by the second drone. 【0124】 <Overview of the anomaly detection system related to Modification 3> The outline of the anomaly detection system according to this modified example 3 will be explained. Figure 14 is an explanatory diagram of the outline of the anomaly detection system according to this modified example 3. 【0125】 In this anomaly detection system, if an object detected based on images taken by the first drone is determined to be in a clearly abnormal state, for example, if flames are detected in a building as shown in Figure 14, the system will notify the monitoring terminal of the anomaly information without performing a determination using images taken by the second drone. 【0126】 Furthermore, if a vehicle or person is detected in a restricted area, it may be determined that this is clearly an abnormal situation, and the abnormal information may be notified to the monitoring terminal without requiring further analysis using the images captured by the second drone. 【0127】 As described above, in the anomaly detection system according to this modified example 3, if the detected object captured by the first drone is determined to be in a clearly abnormal state, the system is configured to notify the monitoring terminal of the anomaly information without performing a determination using the image from the second drone. This makes it possible to efficiently achieve both rapid and wide-ranging information gathering about the monitoring area and reliable detection of abnormal conditions. 【0128】 [Differentiation Example 4] By the way, in the above embodiment, when a first drone patrolling and monitoring at high altitude determines that there is an anomaly based on the image it has captured, a second drone patrolling and monitoring at low altitude circles and monitors the area above the location where the anomaly was determined to be present, and if the second drone determines that there is an anomaly based on the image it has captured, it notifies the monitoring officer's terminal of the anomaly information. However, the present invention is not limited to this. 【0129】 When the first drone detects an anomaly based on images it has captured, the second drone can be configured to use its own lighting to perform low-altitude monitoring of the location. If the illumination required for imaging is insufficient, such as at night. 【0130】 In this modified example 4, we describe an anomaly detection system in which, when the second drone circles at a low altitude to monitor a location where an anomaly has been determined based on images taken by the first drone, if there is insufficient illumination for imaging, such as at night, the second drone uses its own lighting to take images. 【0131】 <Overview of the anomaly detection system related to Modification 4> The outline of the anomaly detection system according to this modified example 4 will be explained. Figure 15 is an explanatory diagram of the outline of the anomaly detection system according to this modified example 4. 【0132】 In this anomaly detection system, as shown in Figure 15(a), when the first drone determines that an anomaly exists based on the images it has captured, the second drone 200 circles at a low altitude to monitor the location. If there is insufficient illumination for imaging, such as at night, the second drone 200 uses its mounted lighting to capture the image. 【0133】 Furthermore, if a person is captured in the image, the system may use AI to detect the person's skeleton and then detect the person's movement, as shown in Figure 15(b). 【0134】 As described above, in the anomaly detection system according to this modified example 4, when the second drone circles at a low altitude to monitor a location where an anomaly has been determined based on the image captured by the first drone, if there is insufficient illumination for imaging, such as at night, the system is configured to use the lighting mounted on the second drone to take images. Therefore, even when there is insufficient illumination for imaging, such as at night, it is possible to efficiently achieve both rapid and wide-area information gathering of the monitoring area and reliable detection of anomalies. 【0135】 [Difference 5] By the way, the above embodiment describes a configuration in which anomalies are detected using a drone when there are no people such as employees in the monitoring area (for example, on holidays or at night), but the present invention is not limited to this. 【0136】 It can also be configured to detect only vehicles within the monitoring area, only people near the boundary of the monitoring area, or only non-employees. 【0137】 This modified example 5 describes an anomaly detection system that detects only vehicles within a monitoring area, only people near the boundary of a monitoring area, or only people who are not employees. 【0138】 <Overview of the anomaly detection system related to Modification 5> The outline of the anomaly detection system according to this modified example 5 will be explained. Figures 16 to 18 are explanatory diagrams illustrating the outline of the anomaly detection system according to this modified example 5. 【0139】 This anomaly detection system detects only vehicles and not people within the monitored area. For example, as shown in Figure 16(a), if a person is visible in the images captured by the first and second drones, it is judged as normal and "no anomaly." As shown in Figure 16(b), if a vehicle is visible in the images captured by the first and second drones, it is judged as "anomaly." This allows for the selection of detection targets to be notified as an anomaly as needed, such as detecting a vehicle entering an area where vehicle entry is prohibited. 【0140】 Furthermore, as shown in Figure 17(a), if the image captured by the drone shows a person walking inside the monitoring area, it is judged as normal and "no abnormality." As shown in Figure 17(b), if a person walking near the boundary of the monitoring area is shown, it is judged as "abnormal." This means that even if a person walking on the premises is detected, it will not be judged as abnormal, but the presence of a person attempting to enter the premises can be notified, thus preventing intruders from entering. 【0141】 Furthermore, as shown in Figure 18(a), if the image captured by the drone shows a person wearing clothing that identifies them as an employee, it is considered normal and judged as "no abnormality." As shown in Figure 18(b), if the image shows a person not wearing employee clothing, it is judged as "abnormal." 【0142】 As described above, the anomaly detection system according to this modified example 5 is configured to detect only vehicles in the monitored area, only people near the boundary of the monitored area, or only people who are not employees. Therefore, it is possible to efficiently achieve both rapid and wide-ranging information gathering about the monitored area and reliable detection of abnormal conditions. 【0143】 It should be noted that the configurations illustrated in the above embodiments and each of the modified examples are functionally schematic and do not necessarily have to be physically arranged as shown. In other words, the distributed and integrated forms of each device are not limited to those shown, and all or part of them can be functionally or physically distributed and integrated in any unit according to various loads and usage conditions. [Industrial applicability] 【0144】 The anomaly detection system, control device, anomaly detection method, and anomaly detection program according to the present invention are suitable for efficiently achieving both the collection of a wide range of information about a monitoring area and the rapid and reliable detection of abnormal conditions. [Explanation of Symbols] 【0145】 N Network 10. First Drone 11 Control device 12 Control Unit 12a Flight path acquisition unit 12b Flight Control Unit 12c Location information acquisition section 12d Image Information Acquisition Unit 12e Information Notification Department 13 Storage section 13a Flight path data 14 Communications Department 15 GNSS receiver 16 cameras 17 Motor 20. Second Drone 21 Control device 22 Control Unit 22a Flight path acquisition unit 22b Flight Control Unit 23 Memory section 23a Flight path data 30 Management device 32 Control Unit 32a First Flight Path Designation Section 32b Second Flight Path Designation Section 32c Drone Control Unit 32d First Image Analysis Unit 32e 1st location information acquisition unit 32f 1st judgment section 32g 2nd image analysis section 32h 2nd location information acquisition unit 32i 2nd judgment section 32j Abnormal notification section 33 Storage section 33a First Flight Path Data 33b Second Flight Path Data 33c Image Data 1 33d Second image data 33e Anomaly Location Data 34 Communications Department 35 Display section 36 Input section 40 Surveillance terminal 110 First Drone 120 Second Drone 200 Second Drone

Claims

[Claim 1] An anomaly detection system comprising a plurality of aircraft equipped with predetermined imaging devices and a control device for controlling the plurality of aircraft, wherein an anomaly is detected using images captured by the plurality of aircraft, The control device is A route designation unit that notifies a first aircraft of a first flight path and notifies a second aircraft of a second flight path at a lower altitude than the first flight path, A first analysis unit detects a target based on a first image captured by an imaging device of a first aircraft flying along the first flight path, If the first analysis unit detects a target, the first determination unit provides at least the second aircraft with a detailed monitoring instruction for the location where the target was detected. An anomaly detection system characterized by having the following features. [Claim 2] The control device is A second determination unit determines whether there is an abnormality based on a second image captured by the imaging device of a second aircraft flying along the second flight path, and if it determines that there is no abnormality, it instructs the second aircraft to return to flight based on the second flight path. The anomaly detection system according to claim 1, further comprising the features described above. [Claim 3] The control device is A second determination unit that determines whether or not there is an abnormality based on a second image captured by an imaging device of a second aircraft flying along the second flight path, If the second determination unit determines that there is an abnormality, the notification unit notifies the abnormality information. The anomaly detection system according to claim 1, further comprising the features described above. [Claim 4] A control device for controlling multiple aircraft equipped with a predetermined imaging device, A route designation unit that notifies a first aircraft of a first flight path and notifies a second aircraft of a second flight path at a lower altitude than the first flight path, A first analysis unit detects a target based on a first image captured by an imaging device of a first aircraft flying along the first flight path, If the first analysis unit detects a target, the first determination unit provides at least the second aircraft with a detailed monitoring instruction for the location where the target was detected. A control device characterized by being equipped with [Claim 5] An anomaly detection method in an anomaly detection system comprising a plurality of aircraft equipped with predetermined imaging devices and a control device for controlling the plurality of aircraft, wherein an anomaly is detected using images captured by the plurality of aircraft, The control device, A route designation step in which a first flight path is notified to a first aircraft, and a second flight path at a lower altitude than the first flight path is notified to a second aircraft, An analysis step to detect a target based on a first image captured by an imaging device of a first aircraft flying along the first flight path, If the detection target is detected by the analysis step, a determination step is performed to give at least the second aircraft a detailed monitoring instruction to the location where the detection target was detected. An anomaly detection method characterized by including [Claim 6] An anomaly detection program executed by a control device that controls multiple aircraft equipped with a predetermined imaging device, A route designation procedure that notifies a first aircraft of a first flight path and notifies a second aircraft of a second flight path at a lower altitude than the first flight path, An analysis procedure for detecting a target based on a first image captured by an imaging device of a first aircraft flying along the first flight path, If the detection target is detected by the analysis procedure, a determination procedure is performed to give at least the second aircraft a detailed monitoring instruction for the location where the detection target was detected. An anomaly detection program characterized by causing a computer to execute a command.

Images