Nest monitoring system and nest monitoring method

Abstract

To efficiently monitor whether or not a bird builds a nest.SOLUTION: A nest building monitor system according to the present invention sores captured image data as data obtained by shooting a site to be monitored of presence or absence of a nest being built by a bird, generates fly history information as information indicating a history of an activity state of a bird from a video of a bird in the captured image, generates, based on the fly history information, a fly history feature amount as a feature amount representing an activity state of the bird from a video of the bird in the captured image data, stores an activity state determination model as a machine learning model obtained by learning, as learning data, data including the fly history feature amount and a label set with respect to the fly history feature amount, gives the activity state determination model the fly history feature amount generated based on the newly acquired fly history information to thereby acquire an activity state determination result as information indicating an activity state of the birds corresponding to the fly history information, and determines whether or not the nest at the site is built based on the activity state determination result to output the determined result.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a nest monitoring system and a nest monitoring method, and particularly to a technique for efficiently monitoring bird nesting in facilities such as substations.

Background Art

[0002] As a bird damage that affects the operation of the power supply system, there is a power outage failure caused by nesting in power-related facilities (substation facilities, power transmission and distribution facilities, etc.). For example, crows use tree branches or wire hangers as materials to build nests on the steel structures of substations, transmission towers, distribution poles, etc. When the nest materials come into contact with wire connection fittings, etc., a short circuit occurs, causing a power outage. Therefore, techniques have been proposed to prevent power outages caused by such bird damage.

[0003] For example, Patent Document 1 describes a nest determination device configured for the purpose of accurately and at low cost detecting the presence of a nest of crows or the possibility of nesting. The nest determination device detects that an animal has contacted or pressed against the detection surface of a sensor attached to a support for supporting an electric wire or a structure provided on the support, counts the number of contacts or the number of presses of the detection surface by the animal based on the detection result, and determines the presence or possibility of nesting by the animal based on the number of contacts or the number of presses counted in the counting step.

[0004] Also, for example, Patent Document 2 describes a nest information management device configured for the purpose of reliably preventing damage to power distribution facilities caused by bird nesting. The nest information management device is connectable via a network to an information provider terminal used by an information provider who reports bird nest information, acquires nest information including nest image information regarding at least the vicinity of a power distribution facility from the information provider terminal, and identifies at least the power distribution facilities related to the nest based on the nest image information and the original power distribution facility image information before the nest corresponding to the nest image information for grasping the nest situation.

Prior Art Documents

[0005] [Patent Document 1] Japanese Patent Publication No. 2015-15867 [Patent Document 2] Japanese Patent Publication No. 2011-204012 [Overview of the Initiative] [Problems that the invention aims to solve]

[0006] For example, substations, a type of power-related facility, contain numerous structures that are complex combinations of steel frames, power equipment, and power lines. Currently, inspectors have to go around to each of these structures and visually check for nesting, which requires a great deal of effort and time. Furthermore, since nesting takes place over a short period, inspections need to be carried out frequently. In addition, since bird nests are often built in high places, inspectors have to look upwards when inspecting the sites, which places a great physical burden on them.

[0007] This invention was made in view of the above background, and aims to provide a nesting monitoring system and a nesting monitoring method that can efficiently monitor whether or not birds are nesting. [Means for solving the problem]

[0008] One of the present inventions for achieving the above objective is a nesting monitoring system, which is configured using an information processing device having a processor and a memory device, and stores photographic data, which is data taken of a site to be monitored for the presence or absence of nesting by birds, generates flight history information, which is information indicating the history of the activity state of birds, from the image of birds captured in the photographic data, generates flight history features, which is information representing the characteristics of the activity state of the birds, based on the flight history information, stores an activity status determination model, which is a machine learning model that has been trained using data including the flight history features based on the flight history information and labels set for the flight history features as training data, and by providing the flight history features generated based on the newly acquired flight history information to the activity status determination model, obtains an activity status determination result, which is information indicating the activity state of birds corresponding to the flight history information, determines the presence or absence of nesting by birds at the site based on the activity status determination result, and outputs the determination result.

[0009] Further issues disclosed in this application, and methods for solving them, will be made clear in the section on embodiments for carrying out the invention and in the drawings. [Effects of the Invention]

[0010] According to the present invention, it is possible to efficiently monitor whether or not birds are nesting. [Brief explanation of the drawing]

[0011] [Figure 1] This diagram shows a schematic configuration of the nesting monitoring system. [Figure 2] This diagram shows the main functions of a nesting monitoring device. [Figure 3] This is an example of the shooting data. [Figure 4A] This is an example of flight history information for birds during nesting season. [Figure 4B] This is an example of flight history information for birds during their resting periods. [Figure 4C] This is an example of flight history information during a bird's feeding time (when it is eating). [Figure 4D] This is an example of the flight history information during bird feeding (feeding chicks). [Figure 5] This is an example of the flight history information. [Figure 6] This is an example of the learning data. [Figure 7] This is a diagram for explaining the activity status determination model. [Figure 8] This is a schematic diagram for explaining the time interval. [Figure 9] This is an example of the determination result presentation screen. [Figure 10] This is a flowchart for explaining the activity status determination model learning process. [Figure 11] This is a flowchart for explaining the activity state determination process. [Figure 12] This is an example of the hardware configuration of the information processing device used to realize the nest building monitoring device.

Mode for Carrying Out the Invention

[0012] Hereinafter, an embodiment will be described with reference to the drawings. In the following description, the character "S" attached before the reference numerals means a processing step.

[0013] FIG. 1 shows a schematic configuration of an information processing system (hereinafter referred to as "nest building monitoring system 1") for monitoring the presence or absence of nest building by birds, which is described as an embodiment of the present invention. As shown in the figure, the nest building monitoring system 1 includes a camera 20 that photographs the periphery of the site where the facility to be monitored (hereinafter referred to as "monitoring target facility") exists, and a nest building monitoring device 100 connected to the camera 20 via a communication network 5. Note that the monitoring target facility is, for example, a power-related facility such as a substation, but the type of the monitoring target facility is not necessarily limited. Also, the bird is, for example, a crow, but the type of the bird is not necessarily limited.

[0014] The camera 20 is a camera (referred to as a "panoramic camera", "360° camera", etc.) that can simultaneously capture the entire sphere (video shooting, still image shooting). The camera 20 records the captured video data or time-series still image data as electronic data in a predetermined data format (hereinafter referred to as "shooting data"), and transmits the recorded shooting data to the nest monitoring device 100 via the communication network 5.

[0015] The communication network 5 is, for example, a wired or wireless communication infrastructure, such as a LAN (Local Area Network), WAN (Wide Area Network), wireless LAN, Internet, 920Hz band communication network, PLC (Power Line Communication), etc.

[0016] The nest monitoring device 100 is configured using an information processing device (computer). The nest monitoring device 100 analyzes the shooting data sent from the camera 20, and provides the user with information regarding the presence or absence of nesting by birds and the possibility of nesting in the monitored facility. The nest monitoring device 100 is provided, for example, in a monitoring station, business office, etc. 2 operated by the management entity of the power-related facility when the monitored facility is a power-related facility. The user (operation entity) of the nest monitoring device 100 is, for example, a manager or inspector of a power-related facility.

[0017] Fig. 2 shows the main functions of the nest monitoring device 100. As shown in the figure, the nest monitoring device 100 includes functions of a storage unit 110, a shooting data acquisition management unit 125, a flight history information generation unit 130, a flight history feature amount calculation unit 135, a learning data generation unit 140, a model learning unit 145, an activity status determination unit 150, a nesting presence / absence determination unit 155, a nesting position specification unit 160, and a determination result presentation unit 165.

[0018] Among the above functions, the storage unit 110 stores information (data) such as shooting data 111, flight history information 112, flight history feature amount 113, learning data 114, activity status determination model 115, activity status determination result 116, nesting presence / absence determination result 117, and nesting position specification result 118.

[0019] The shooting data acquisition and management unit 125 receives the shooting data sent from the camera 20 and manages the received shooting data as shooting data 111.

[0020] The flight history information generation unit 130 acquires information from the photographic data 111 that shows the bird's activity (bird's flight trajectory, how the bird is perched on a structure, how the bird is walking, etc.) and manages the acquired information as flight history information 112.

[0021] Figure 3 shows an example of the captured data 111. The flight history information generation unit 130 detects birds captured in the captured data 111 using, for example, a known object detection mechanism (R-CNN (Regional CNN (Convolutional Neural Network)), YOLO (You only Look once), SDD (Single Shot Detector), DETR (DEtection Transformer), etc.), and generates flight history information 112 based on, for example, information indicating the position of the birds captured in each frame of the captured data (i.e., time-series data of information indicating the position of the birds).

[0022] Figures 4A to 4D schematically illustrate examples of bird arrival history information 112 for different bird activity situations. Of these, Figure 4A is an example of bird arrival history information 112 during nesting. Figure 4B is an example of bird arrival history information 112 during resting. Figure 4C is an example of bird arrival history information 112 during feeding (when the bird is eating). Figure 4D is an example of bird arrival history information 112 during feeding (when the bird is feeding its chicks). The horizontal axis of the graphs in each of these figures represents time. Each band in the graphs in each figure corresponds to a single bird arrival (from the start to the end of its appearance in the photographic data). The color of the band indicates whether the bird is in flight (white in the figure) or stationary (black in the figure).

[0023] As shown in Figure 4A, during nesting, birds repeatedly fly to the location of the nest at short intervals. Also, the time that birds spend at the nest (hereinafter referred to as "staying time") is often short.

[0024] As shown in Figure 4B, during rest periods, birds fly in at longer intervals than during nesting. Furthermore, the duration of rest periods tends to be longer than during nesting periods.

[0025] As shown in Figure 4C, during feeding time, birds fly in several times in quick succession until they finish eating their food. However, the number of flights per given time is less than during nesting. Also, the resting time is longer than during nesting but shorter than during resting.

[0026] As shown in Figure 4D, during feeding, birds fly in repeatedly, but because it takes time to obtain food, the time intervals between visits are longer than during nesting. Also, since they fly away immediately after giving food to their chicks, the time birds spend at the nest (stopping time) is often short.

[0027] Thus, the activities of birds (nesting, resting, eating, feeding) can be characterized by the number of consecutive visits (hereinafter referred to as "number of consecutive visits"), the time interval between visits, and the resting time (hereinafter, these three features are referred to as "visit history features"). It should be noted that the types of bird activities are not necessarily limited to those shown in Figures 4A to 4D (nesting, resting, eating, feeding).

[0028] Figure 5 shows an example of the flight history information 112. As shown in the figure, the example flight history information 112 consists of multiple records, each containing the following items: flight time (before stopping) 511, stopping time 512, flight time (after stopping) 513, elapsed time since the last visit 514, stopping position 515, and arrival date and time 516. One record of the flight history information 112 corresponds to one visit by a bird.

[0029] Of the above items, Flight Time (Before Stopping) 511 stores the time from when the bird arrives until it stops flying (lands) (from when it starts appearing in the photographic data until it stops flying). Stop Time 512 stores the time the bird is stopped flying. Flight Time (After Stopping) 513 stores the time from when the bird, which had stopped flying, starts flying again until it finishes appearing in the photographic data. Elapsed Time Since Previous Arrival 514 stores the time from when the bird last arrived (when it finished appearing in the previous photographic data) to when it arrived this time (when it started appearing in the current photographic data). Stopping Position 515 stores information indicating the position of the bird when it stopped flying. Note that the above position is represented, for example, by the position coordinates (two-dimensional or three-dimensional coordinates) of the coordinate system set for the photographic data. Arrival Date and Time 516 stores the date and time the bird arrived (the date and time it started appearing in the current photographic data).

[0030] Returning to Figure 2, the flight history feature calculation unit 135 calculates flight history features (number of consecutive flights, flight time interval, stop time) based on the flight history information 112, and manages the calculated flight history features as flight history features 113.

[0031] The training data generation unit 140 receives the user's setting of labels (ground truth data) for each of the flight history features (flight history features based on flight history information acquired in the past) managed as flight history features 113, and generates data that associates the flight history features with the accepted labels as training data 114 to be used for training the activity status determination model 115. The activity status determination model 115 is a machine learning model (e.g., DNN (Deep Neural Network), decision tree, support vector machine, etc.) that determines the status of bird activity based on flight history features. The activity status determination model 115 can be represented, for example, by a matrix, mathematical formula, vector, etc., which include adjustable parameters.

[0032] Figure 6 shows an example of training data 114. As shown in the figure, the example training data 114 consists of multiple records, each having the following items: number of consecutive flights 611, flight time interval 612, stop time 613, and label 614. One record of training data 114 corresponds to one of the training data.

[0033] Of the above items, the number of consecutive visits 611, the visit time interval 612, and the stop time 613 constitute explanatory variables input to the activity status determination model 115. The number of consecutive visits 611 stores the number of consecutive visits. The visit time interval 612 stores the average value (seconds) of the visit time interval. The stop time 613 stores the average value (seconds) of the stop time. The label 614 stores the label (ground truth data, target variable) for the explanatory variable.

[0034] Returning to Figure 2, the model learning unit 145 trains the activity status determination model 115 based on the training data 114.

[0035] Figure 7 shows an example of the activity status determination model 115 (a diagram showing the structure of a neural network). As shown in the figure, the input layer 711 of the activity status determination model 115 is input with the explanatory variables (flight history features) mentioned above. The hidden layer 712 contains one or more hidden layers consisting of one or more nodes containing parameters that are adjusted by learning. The hidden layer 712 calculates one or more predicted values ​​(probabilities) for the output layer 713 based on the flight history features given to the input layer 711. The output layer 713 stores the probabilities for each type of bird activity (nesting, resting, eating, feeding) (probability of nesting, probability of resting, probability of eating, probability of feeding).

[0036] Returning to Figure 2, the activity status determination unit 150 generates a flight history feature 113 based on the flight history information 112 for each predetermined time interval of the photographic data 111 acquired at the site, and inputs the generated flight history feature 113 as an explanatory variable into the activity status determination model 115 to obtain the activity status of birds.

[0037] Figure 8 is a schematic diagram illustrating the above time intervals. The activity status determination unit 150 defines each period with a predetermined time width TM from the start time as the above time interval, which is obtained sequentially by shifting the start time by, for example, a predetermined time shift amount Δt. For example, the nesting monitoring device 100 may provide a user interface for setting the time shift amount Δt and the time width TM. By appropriately setting the time shift amount Δt and the time width TM according to, for example, the type of bird, the site conditions, the season, etc., the accuracy of determining the activity status of birds can be improved.

[0038] Returning to Figure 2, the nesting presence / absence determination unit 155 determines whether or not the bird has built a nest based on the bird's activity status determined by the activity status determination unit 150, and stores information indicating the determination result in the nesting presence / absence determination result 117. For example, the nesting presence / absence determination unit 155 determines that the bird has built a nest when the activity status determination unit 150 determines the bird's activity status to be "nesting" or "feeding" (i.e., the probability of "nesting" or the probability of "feeding" exceeds a preset threshold). The nesting monitoring device 100 may be provided with a user interface for setting the above thresholds. By appropriately setting the above thresholds according to differences such as the type of bird or the season, the accuracy of determining the bird's activity status can be improved.

[0039] The nesting location identification unit 160 identifies the nesting location based on the information of the stopping position 515 in the flight history information 112 used for the determination when the nesting presence / absence determination unit 155 determines that a bird has built a nest, and stores the information indicating the identified location in the nesting location identification result 118. For example, the nesting location identification unit 160 identifies the location of the nest as the position where the endpoints of multiple flight paths identified from the flight history information 112 converge.

[0040] The result presentation unit 165 presents the activity status determination result 116, the nesting presence / absence determination result 117, and the nesting location identification result 118 to the user via the user interface.

[0041] Figure 9 shows an example of a screen presented to the user by the judgment result presentation unit 165 (hereinafter referred to as the "judgment result presentation screen 900"). As shown in the figure, the example judgment result presentation screen 900 has a display period specification field 911, a display button 912, a flight path display field 913, a nest presence / absence judgment result display field 914, and an activity status judgment result display field 915.

[0042] The display period specification field 911 is used to specify the period for which the user wants to check the judgment results. When the user operates the display button 912, an image showing the bird's flight path (activity status) during the period specified in the display period specification field 911 (an image generated by the judgment result presentation unit 165 based on the flight history information 112) is displayed in the flight path display field 913. In the figure, the dashed lines indicate the bird's flight path. Furthermore, if the bird is nesting or feeding, the endpoints of multiple flight paths will be at the same position (the position of the "○" shown in the figure), allowing the user to identify (confirm) a location where a nest is likely to exist.

[0043] Furthermore, when the user operates the display button 912, the nest presence / absence determination result display field 914 displays the contents of the nest presence / absence determination result 117 for the above period. In addition, the activity status determination result display field 915 displays the contents of the activity status determination result 116 (probability for each type of activity status) for the relevant period.

[0044] Figure 10 is a flowchart illustrating the processes (hereinafter referred to as "activity status determination model learning process S1000") that the nesting monitoring device 100 performs when generating training data 114 and training the activity status determination model 115 using the generated training data 114. The timing of executing the activity status determination model learning process S1000 is not necessarily limited, but for example, the nesting monitoring device 100 may execute the activity status determination model learning process S1000 when a new flight history feature 113 is generated or when it receives an execution instruction from the user via the user interface. The activity status determination model learning process S1000 will be explained below with reference to the same figure.

[0045] First, the training data generation unit 140 presents the flight history features (explanatory variables) to the user, accepts the user's setting of labels (target variables) for said flight history features, and generates training data 114 by associating said flight history features with the labels (S1011).

[0046] Next, the model learning unit 145 learns the activity status determination model 115 based on the training data 114 (S1012). Alternatively, the model learning unit 145 may perform, for example, a verification of the prediction accuracy of the trained activity status determination model 115. In that case, for example, the training data 114 may be pre-classified into training data and validation data, with the training data used for training and the validation data used for verification.

[0047] Figure 11 is a flowchart illustrating the process by which the nesting monitoring device 100 inputs explanatory variables to be judged (for example, newly acquired flight history features in a time interval) into the activity status determination model 115 to obtain the target variable (probability of nesting, probability of resting, probability of feeding, probability of feeding), and presents the obtained target variable to the user (hereinafter referred to as "activity status determination process S1100"). The activity status determination process S1100 will be explained below with reference to the same figure.

[0048] First, the flight history information generation unit 130 generates new flight history information 112 based on the captured data 111 (S1111).

[0049] Next, the flight history feature calculation unit 135 generates flight history feature quantities 113 based on the new flight history information 112 (S1112).

[0050] Next, the activity status determination unit 150 inputs the flight history features 113 (explanatory variables) into the activity status determination model 115 to calculate the target variables (probability of nesting, probability of resting, probability of eating, probability of feeding) and generates the activity status determination result 116 (S1113).

[0051] Next, the nesting presence / absence determination unit 155 determines whether or not a nest is present based on the activity status determination result 116 and generates a nesting presence / absence determination result 117 that stores the determination result (S1114).

[0052] Next, the judgment result presentation unit 165 generates and presents the judgment result presentation screen 900 to the user, receives the display period from the user, and displays the contents of the flight path display field 913, the nesting presence / absence judgment result display field 914, and the activity status judgment result display field 915 based on the received information (S1115).

[0053] As described in detail above, the nesting monitoring system 1 of this embodiment automatically determines whether or not birds are nesting at the site by obtaining information indicating the activity status of birds corresponding to the flight history information by providing flight history features generated based on newly acquired flight history information to the activity status determination model. Therefore, the burden of monitoring for the presence or absence of nesting at the site can be reduced. In particular, in places where there are many complex structures such as substations, it is not necessary for people to look up and patrol the premises one by one, and the presence or absence of nesting can be monitored efficiently.

[0054] Furthermore, the nesting monitoring system uses an activity status determination model that outputs a target variable with parameters suitable for determining bird activity, such as the number of consecutive visits, the time interval between visits, and the resting time, as explanatory variables, to determine whether or not birds are nesting. Therefore, it can accurately determine whether or not birds are nesting.

[0055] Furthermore, since the captured data is obtained by a 360-degree camera installed on site and captures a wide area of ​​the site, it is possible to monitor bird activity over a wide area with minimal equipment, and to efficiently monitor for the presence or absence of nesting.

[0056] <Example of an information processing device> Figure 12 shows an example of the hardware configuration of an information processing device used to realize the nesting monitoring device 100. The illustrated information processing device 10 comprises a processor 11, a main memory 12, an auxiliary memory 13, an input device 14, an output device 15, and a communication device 16. Specific examples of the information processing device 10 include, for example, a personal computer, an office computer, various server devices, and a general-purpose computer. The information processing device 10 may be implemented, in whole or in part, using virtual information processing resources provided using virtualization technology or process space isolation technology, such as a virtual server provided by a cloud system. The nesting monitoring device 100 may also be implemented using multiple information processing devices 10 that are connected to each other in a communicative manner.

[0057] In the figure, the processor 11 is composed of, for example, a CPU (Central Processing Unit), an MPU (Micro Processing Unit), a GPU (Graphics Processing Unit), an FPGA (Field Programmable Gate Array), an ASIC (Application Specific Integrated Circuit), an AI (Artificial Intelligence) chip, etc.

[0058] The main memory 12 is a device for storing programs and data, and is, for example, ROM (Read Only Memory), RAM (Random Access Memory), or non-volatile memory (NVRAM (Non-Volatile RAM)).

[0059] The auxiliary storage device 13 includes, for example, an SSD (Solid State Drive), a hard disk drive, an optical storage device (CD (Compact Disc), DVD (Digital Versatile Disc), etc.), a storage system, an IC card, a reader / writer for recording media such as SD cards and optical recording media, and the storage area of ​​a cloud server. Programs and data can be read into the auxiliary storage device 13 via a recording media reader or a communication device 16. Programs and data stored in the auxiliary storage device 13 are read into the main memory 12 as needed.

[0060] The input device 14 is an interface that accepts input from an external source, and can be, for example, a keyboard, mouse, touch panel, card reader, pen-input tablet, or voice input device.

[0061] The output device 15 is an interface that outputs various information such as processing progress and processing results. The output device 15 may be, for example, a display device that visualizes the above information (LCD (Liquid Crystal Display), graphics card, etc.), a device that converts the above information into sound (speaker, etc.), or a device that converts the above information into text (printer, etc.). For example, the information processing device 10 may be configured to input and output information to and from other devices via the communication device 16.

[0062] The input device 14 and the output device 15 constitute a user interface for receiving and presenting information with the user.

[0063] The communication device 16 is a device that enables communication (wired or wireless communication) with other devices via a communication infrastructure such as the communication network 5, and is configured using, for example, a NIC (Network Interface Card), a wireless communication module, a USB module, etc.

[0064] The information processing device 10 may have, for example, an operating system, a file system, a DBMS (Database Management System) (relational database, NoSQL, etc.), a KVS (Key-Value Store), etc. installed on it.

[0065] The functions of the nest monitoring device 100 are realized either by the processor 11 of the information processing device 10 reading and executing a program stored in the main memory 12, or by the functions of the hardware (FPGA, ASIC, AI chip, etc.) that constitutes the nest monitoring device 100 itself. The nest monitoring device 100 stores the aforementioned various types of information (data) as, for example, database tables or files managed by a file system.

[0066] The embodiments of the present invention have been described in detail above, but this description is for the purpose of facilitating understanding of the present invention and does not limit it. The present invention can be modified and improved without departing from its spirit, and of course, equivalents thereof are included in the present invention. For example, the above embodiments have been described in detail for the purpose of explaining the present invention in an easy-to-understand manner and are not necessarily limited to those having all the configurations described. Furthermore, it is possible to add, delete, or replace some of the configurations of the above embodiments with other configurations.

[0067] For example, the accuracy of determining the activity status can be improved by using other types of features obtained from flight history information generated from the captured data as explanatory variables. Alternatively, if the user-required accuracy in determining the activity status can be ensured, one or two of the aforementioned features (number of consecutive flights, flight time interval, stop time) can be selected and used as explanatory variables. [Explanation of Symbols]

[0068] 1 Nesting monitoring system, 2 Monitoring station / office, etc., 5 Communication network, 100 Nesting monitoring device, 110 Memory unit, 111 Photographed data, 112 Flight history information, 113 Flight history features, 114 Training data, 115 Activity status determination model, 116 Activity status determination result, 117 Nest presence / absence determination result, 118 Nest location identification result, 125 Photographed data acquisition management unit, 130 Flight history information generation unit, 135 Flight history features calculation unit, 140 Training data generation unit, 145 Model training unit, 150 Activity status determination unit, 155 Nest presence / absence determination unit, 160 Nest location identification unit, 165 Judgment result presentation unit

Claims

1. It is configured using an information processing device having a processor and a memory device, The system stores photographic data, which is data taken from locations that are being monitored for the presence or absence of bird nesting. From the images of birds captured in the aforementioned photographic data, information indicating the history of the birds' activity is generated, including their arrival history. Based on the aforementioned flight history information, a flight history feature quantity is generated, which is information that represents the activity state of the bird. The machine learning model, which is an activity status determination model, is trained using data including the flight history features based on the aforementioned flight history information and the labels set for said flight history features as training data, and stores this model. By providing the flight history features generated based on the newly acquired flight history information to the activity status determination model, an activity status determination result is obtained, which is information indicating the activity status of the bird corresponding to the flight history information. Based on the activity status determination results, the presence or absence of nesting by birds at the site is determined, and the determination result is output. Nest monitoring system.

2. A nesting monitoring system according to claim 1, The aforementioned flight history feature includes the number of consecutive flights, which is the number of times a bird has flown in consecutively; the flight time interval, which is the time interval between bird flights; and the resting time, which is the time a bird is stationary. Nest monitoring system.

3. A nesting monitoring system according to claim 2, The aforementioned number of consecutive arrivals is the number of times that arrivals occur consecutively at time intervals below a predetermined threshold. Nest monitoring system.

4. A nesting monitoring system according to claim 3, Having a user interface for setting the aforementioned threshold, Nest monitoring system.

5. A nesting monitoring system according to claim 1, The activity state of the aforementioned bird is at least one of the following: nesting, resting, eating, and feeding. Nest monitoring system.

6. A nesting monitoring system according to claim 5, If the probability of nesting or feeding, as shown in the activity status determination results, exceeds a predetermined threshold, it is determined that nesting by birds has occurred at the site. Nest monitoring system.

7. A nesting monitoring system according to claim 6, Having a user interface for setting the aforementioned threshold, Nest monitoring system.

8. A nesting monitoring system according to claim 1, The aforementioned photographic data was acquired by a 360-degree camera installed at the site. Nest monitoring system.

9. A nesting monitoring system according to claim 1, The aforementioned flight history information is generated by performing object detection processing on the aforementioned photographic data. Nest monitoring system.

10. An information processing device having a processor and a memory device, The step of storing photographic data, which is data taken from a site that is being monitored for the presence or absence of bird nesting, A step of generating flight history information, which is information indicating the history of the bird's activity state, from the image of the bird captured in the aforementioned shooting data. A step of generating a flight history feature quantity, which is information that represents the activity state of the bird, based on the aforementioned flight history information. A step of storing an activity status determination model, which is a machine learning model that has been trained using data including the flight history features based on the flight history information and the labels set for the flight history features as training data. The steps include: obtaining an activity status determination result, which is information indicating the activity status of a bird corresponding to the flight history information, by providing the flight history feature quantities generated based on the newly acquired flight history information to the activity status determination model; and A step of determining whether or not birds are nesting at the site based on the activity status determination result, and outputting the determination result. A method for monitoring nests.

11. A nesting monitoring method according to claim 10, The aforementioned flight history feature includes the number of consecutive flights, which is the number of times a bird has flown in consecutively; the flight time interval, which is the time interval between bird flights; and the resting time, which is the time a bird is stationary. Nest monitoring methods.

12. A nesting monitoring method according to claim 10, The aforementioned photographic data was acquired by a 360-degree camera installed at the site. Nest monitoring methods.

Images