Nesting response work support device and nesting response work support method

Abstract

To provide information on what and when to do about a nest built on a power facility.SOLUTION: A nest coping work support device 1 includes: a processor; a memory; an image acquisition unit 11 for acquiring an image; an emergency level determination unit 17 for outputting information on the content and the timing of coping with a nest in an acquired image, by inputting the image into a learned model in which an image including a power facility and a nest of birds is input and information on the content and the timing of coping with the nest is output.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a nesting response work support device and a nesting response work support method. [Background technology]

[0002] In recent years, particularly in urban areas, there has been an increase in the number of crows and other birds nesting on utility poles using wire hangers and other materials instead of twigs. These nests may come into contact with power lines and other electrical equipment, potentially causing power outages. For this reason, power company distribution employees and employees of subcontracted companies patrol the equipment to check for nesting and take measures such as removing the nests. However, it is extremely difficult to thoroughly inspect and understand the numerous pieces of equipment installed over a wide area with limited personnel.

[0003] Therefore, as a technology for automating part of the nest inspection, for example, Patent Document 1 discloses that a nest information management server acquires nest image information from an information provider terminal, identifies equipment related to nesting based on the nest image and the original image, determines the urgency of the response, determines whether or not countermeasures are required, and if countermeasures are required, determines the content of the countermeasure work, and displays the results of the urgency determination process and the countermeasure work content determination process on a display unit. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2011-204012 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the criteria for determining nesting in Patent Document 1 require predefined information, such as whether a lightning arrester is buried in the nest, whether a nest is built on an open-type switchgear or a bushing, or whether a cutout is buried in the nest. Because the impact of nesting on each piece of power equipment varies depending on, for example, the structure of the utility pole and the type of associated power equipment, applying the technology of the above Patent Document to a large number of utility poles installed over a wide area may result in complicated definitions of the criteria and may be impractical. Furthermore, even if the criteria for determining whether countermeasures are implemented are established, the actual determination of when and what measures to take against nesting is based on the long-term experience and know-how of on-site workers, etc., and may not be amenable to standardization like that of the above Patent Document.

[0006] The present invention has been made in consideration of this background, and its purpose is to provide a nesting response work support device and a nesting response work support method that are capable of providing information regarding the response that should be taken in response to nests built in power equipment and the timing of that response. [Means for solving the problem]

[0007] One aspect of the present invention for achieving the above-mentioned object is a nest response work support device that has a processor and memory, and is equipped with an image acquisition unit that acquires images, and an urgency determination unit that inputs images including power equipment and bird nests into a trained model that inputs the acquired images and outputs information on the response content and timing of response to the bird nests, and outputs the acquired images to the trained model that outputs information on the response content and timing of response to the bird nests.

[0008] In this way, the nesting response work support device of the present invention inputs site images into a nesting risk determination model, which is a trained model that outputs information indicating the likelihood of future bird nests approaching power equipment, and outputs information on the response content and timing for the nests captured in the site images, so that workers, etc. can know the specific response to be taken and the timing for that response just by preparing an image of a utility pole, etc. where a nest is being built. In other words, the nesting response work support device of the present invention can provide information on the response to be taken and the timing for that response to nests built on power equipment.

[0009] Furthermore, one aspect of the present invention for achieving the above-mentioned object is that the trained model is a model to which information on the area and time when the image was taken is further input, and the urgency determination unit further inputs information on the area and time when the acquired image was taken into the trained model, thereby outputting information on the response content and timing of response to the nest contained in the acquired image.

[0010] In this way, the nesting response work support device according to the present invention outputs information on the response content and timing for the nest contained in the on-site image by inputting information on the area and time when the on-site image was taken into a trained model (nesting risk assessment model) to which information on the area and time when the image was taken is input. Because the nesting rate of birds can vary regionally or seasonally, by inputting the area and time when the on-site image was taken as feature quantities of the nesting risk assessment model, workers and others can accurately know the specific response that should be taken in response to nesting.

[0011] Furthermore, one of the present inventions for achieving the above-mentioned object is a nest response work support device, wherein the trained model is a model that outputs, as information on the response content and response timing for the bird's nest, either information indicating that the nest needs to be removed immediately, information indicating that the nest needs to be removed by a specified time in the future, or information indicating that removal of the nest is not necessary, and the urgency determination unit inputs the acquired image into the trained model, and outputs, as information on the response content and response timing, either information indicating that the nest contained in the acquired image needs to be removed immediately, information indicating that the nest needs to be removed by a specified time in the future, or information indicating that removal of the nest is not necessary.

[0012] In this way, the nesting response work support device according to the present invention inputs a site image into the nesting risk assessment model, and outputs information indicating that the nest included in the site image needs to be removed immediately, information indicating that the nest needs to be removed by a predetermined time in the future, or information indicating that the nest does not need to be removed. This allows workers and others to know more specific measures to take in response to nesting.

[0013] In addition, one aspect of the present invention for achieving the above-mentioned object is a nest response work support device that is equipped with a display unit that displays on a screen information regarding the response content and response timing for the nest contained in the acquired image.

[0014] In this way, the nest response work support device of the present invention displays on the screen information on the response content and timing of the response to the nest captured in the on-site image, so that workers and others can quickly understand the response that should be taken in response to the nest.

[0015] Furthermore, one of the present inventions for achieving the above-mentioned object is a nest response work support device, in which the urgency determination unit inputs an image including power equipment and bird nests into a first trained model that outputs information about the areas of the bird nests and power equipment in the image, thereby outputting information about the areas of the bird nests and power equipment in the acquired image, and inputs information about the areas of the bird nests and power equipment into a second trained model that inputs information about the areas of the bird nests and power equipment and outputs information about the response content and timing for the bird nests, thereby outputting information about the response content and timing for the nests included in the acquired image.

[0016] In this way, the nesting response work support device according to the present invention outputs information on response and timing using a trained model (nesting risk determination model) consisting of a first trained model and a second trained model. That is, the nesting response work support device 1 according to the present invention inputs an image including a power facility and a bird's nest into the first trained model, which outputs information on the areas of the bird's nest and the power facility in the image, thereby outputting information on the areas of the bird's nest and the power facility in the site image. Next, the information on the areas of the bird's nest and the power facility output by the first trained model is input into the second trained model, which outputs information on the response content and timing for the bird's nest, thereby outputting information on the response content and timing for the nest included in the site image. In this way, by using the second trained model with the information on the areas of the nest and the power facility recognized by the first trained model as input values, the response content and timing for the bird's nest can be more accurately estimated.

[0017] Furthermore, one aspect of the present invention for achieving the above-mentioned object is a nest response work support method, in which an information processing device executes an image acquisition process for acquiring an image, and an urgency assessment process for outputting information on the response content and timing of response to the nest contained in the acquired image by inputting the acquired image into a trained model that inputs an image including power equipment and a bird's nest and outputs information on the response content and timing of response to the bird's nest. [Effects of the Invention]

[0018] According to the present invention, it is possible to provide information on what action should be taken against nests that have been built in power facilities and the timing of such action. [Brief explanation of the drawings]

[0019] [Figure 1] FIG. 10 is a diagram illustrating an example of a performance image. [Figure 2] FIG. 10 is a diagram illustrating another example of the performance image. [Figure 3] FIG. 2 is a diagram illustrating an example of functions provided in the nest response operation support device. [Figure 4] FIG. 10 is a diagram illustrating an example of a teacher data DB. [Figure 5] FIG. 2 is a diagram illustrating an example of hardware of a nest response operation support device. [Figure 6] FIG. 10 is a flow diagram illustrating an example of a model learning process. [Figure 7] FIG. 10 is a flow diagram illustrating an example of an urgency determination process. [Figure 8] 10 is an example of an urgency determination screen. DETAILED DESCRIPTION OF THE INVENTION

[0020] Hereinafter, a nest response operation support device according to one embodiment of the present invention will be described with reference to the drawings. On utility poles installed in power systems, birds such as crows and sparrows sometimes build nests using metal parts such as wires or hangers, using the cross arms or electric wires attached to the utility poles as footholds. If nests built in this way are left unattended, the nests may come into contact with surrounding power equipment and cause serious accidents such as short circuits.

[0021] Therefore, the nesting response work support device of this embodiment performs predetermined machine learning using images of bird nests and nearby power equipment (hereinafter referred to as actual images) taken in the past by workers or the like as training data, to create a trained model (hereinafter referred to as a nesting risk determination model) that outputs information on measures to be taken against nests.After that, when an image (hereinafter referred to as an on-site image) taken by a worker or the like who newly discovers nesting on a power equipment such as a utility pole is input into the nesting risk determination model, the nesting response work support device outputs information on the response that the worker or the like should take against the nest and the timing, and displays it on the screen.

[0022] (Actual image) 1 is a diagram showing an example of an actual image 50. This actual image 50 is an image of a utility pole 51 erected from the ground surface (not shown) taken at a predetermined elevation angle from a position a predetermined distance away.

[0023] Specifically, utility pole 51 includes first cross arm 54, which extends horizontally from a predetermined height of support 52 and has switch 53 attached to its tip, and second cross arm 56, which extends horizontally from a predetermined height of support 52 above first cross arm 54. Multiple high-voltage electrical wires 55 are installed on second cross arm 56. Bird nests 59 are made using metal members (not shown), such as wires or hangers, carried by birds from elsewhere, using first cross arm 54, second cross arm 56, or high-voltage electrical wire 55 as footholds. If such nests 59 grow large, they may come into contact with power equipment near first cross arm 54 or second cross arm 56, such as switch 53 or high-voltage electrical wire 55.

[0024] 2 is a diagram showing another example of an actual image 70. Similar to the previously described actual image 50, this actual image 70 is an image of a utility pole 71 erected from the ground (not shown) taken at a predetermined elevation angle from a position a predetermined distance away.

[0025] Specifically, a pole-mounted transformer 73 is mounted on a support 72 of this utility pole 71. The utility pole 71 also includes a first cross arm 76 that extends horizontally from a predetermined height of the support 72 and has a high-voltage cutout support arm 75 with a high-voltage cutout 74 fixed to its tip, and a second cross arm 77 that extends horizontally from the support 72 above the first cross arm 76. A plurality of lightning arresters 78 and high-voltage wires 79 are mounted on the tip of the second cross arm 77. Bird nests 80 are made using metal members (not shown) such as wires or hangers that have been carried by the birds from elsewhere, using the first cross arm 76, the second cross arm 77, or the high-voltage wires 79 as footholds. If such a nest 80 grows large, it may come into contact with power equipment near the first cross arm 76 and the second cross arm 77, such as the high-voltage cutout 74, the high-voltage cutout support arm 75, the lightning arrester 78, or the high-voltage power line 79.

[0026] In this way, as the nests 59 and 80 grow and the distance between them and the surrounding electric power equipment becomes shorter, the possibility that the nests 59 and 80 will come into contact with the electric power equipment increases.

[0027] Furthermore, the growth rate of the nests 59, 80 depends on the activity level of the birds, i.e., the time of year (season, month, etc.) when the nests 59, 80 are built. The growth rate of the nests 59, 80 also depends on the region in which the birds live. Note that differences in the region include not only simple differences in location, but also differences in land use, such as whether the region is an urban area or a mountainous area.

[0028] From the above, it is thought that the urgency of measures such as removing nests 59 and 80 is correlated with the timing and region of nesting, as well as the relative positions of nests 59 and 80 and each power facility. Therefore, the nesting response work support device of this embodiment utilizes such correlations to generate a nesting risk assessment model based on actual images.

[0029] The nest response support device will be described in detail below. --Connected phase determination device-- 3 is a diagram illustrating an example of functions of the nesting response work support device 1 according to this embodiment. The nesting response work support device 1 is an information processing device (computer) including an image acquisition unit 11, a preprocessing unit 13, a learning unit 15, an urgency determination unit 17, and an information display unit 19.

[0030] The nest response work support device 1 also includes a teacher data DB200 that stores data on feature amounts and correct labels of past images, and an image DB300 that stores past images and on-site images.

[0031] The image acquisition unit 11 acquires images (actual images and on-site images) by reading them from a recording medium or receiving them via a communication network.

[0032] Based on the teacher data DB200, the learning unit 15 inputs an image (site image) including power equipment and bird nests, and creates a trained model (nesting risk assessment model) that outputs information on the response content and timing of the response to the bird nests (hereinafter referred to as urgency).

[0033] The trained model may be a model to which information on the area and time when the site image was taken is further input. The nest risk determination model of this embodiment is a model to which information on the area and time when the site image was taken is further input.

[0034] This trained model is constructed by machine learning the feature quantities of on-site images using deep learning. In this embodiment, this trained model is a neural network that has an input layer to which pixel information of the image (pixel content and coordinates) is input, one or more intermediate layers (hidden layers) that extract and output image feature quantities from the pixel information, and an output layer that outputs the urgency level from the image feature quantities.

[0035] As the neural network in the trained model of this embodiment, for example, a convolution neural network (CNN), a support vector machine (SVM), a Bayesian network, or a regression tree can be applied, but this embodiment is based on CNN, which is a method suitable for image recognition.

[0036] The urgency determination unit 17 inputs the on-site image acquired by the image acquisition unit 11 into the trained model, and outputs information on the response content and response timing corresponding to the nest contained in the on-site image.

[0037] The information display unit 19 displays the information output by the urgency determination unit 17 on the screen.

[0038] (Teacher data DB) 4 is a diagram showing an example of a training data DB 200. The training data DB 200 stores actual images and related information. The training data DB 200 includes the following data items: utility pole ID 201, which sets the ID of the utility pole in each image; shooting date and time 202, which sets the shooting date and time of each image; image data 203, which sets information indicating the data file in which each image is recorded; shooting location 204, which sets the shooting location of each image; nest area 205, which sets area information of the nest in each image (information about the content and location of each pixel in the image); power equipment area 206, which sets area information of the power equipment in each image (information about the content and location of each pixel in the image); power equipment type 207, which indicates the type of power equipment in each image; and urgency 208 for nest removal in each image.

[0039] The power equipment type 207 is set with information on the type of power equipment, such as each switch, lightning arrester, transformer, high voltage cutout, high voltage arm, and high voltage cutout support arm, and the power equipment area 206 is set with area information in the actual image.

[0040] In this embodiment, the urgency level 208 is set to one of the following: information indicating that the nest needs to be removed immediately, information indicating that the nest needs to be removed within three days, information indicating that the nest needs to be removed within one week, or information indicating that the nest does not need to be removed. Note that the number of days is not intended to be limited to this value.

[0041] Each data item in the training data DB 200 is set by, for example, a worker involved in monitoring or removing nests while referring to the actual image. Note that the photographing date and time 202 and the photographing location 204 may be extracted from the metadata of the actual image, or may be set by the worker who photographed the actual image.

[0042] 4 are merely examples. In addition to these data items, information that influences the response to nesting and the timing of that response (for example, bird species) may be set as a feature of the nesting risk determination model.

[0043] 5 is a diagram showing an example of hardware of the nesting response work support device 1. The nesting response work support device 1 includes a processor 31, a main memory device 32, an auxiliary memory device 33, an input device 34, an output device 35, and a communication device 36.

[0044] The processor 31 is configured using, 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.

[0045] The main memory device 32 is a device that stores programs and data, and is, for example, a read-only memory (ROM), a random access memory (RAM), or a non-volatile memory (NVRAM (Non-Volatile RAM)).

[0046] The auxiliary storage device 33 is, 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, a read / write device for a recording medium such as an IC card, an SD card or an optical recording medium, a storage area of a cloud server, etc. Programs and data can be read into the auxiliary storage device 33 via a recording medium reader or a communication device 36. The programs and data stored (memorized) in the auxiliary storage device 33 are read into the main storage device 32 as needed.

[0047] The input device 34 is an interface that accepts input from the outside, and is, for example, a keyboard, a mouse, a touch panel, a card reader, a pen-input tablet, a voice input device, or the like.

[0048] The output device 35 is an interface that outputs various information such as the progress of processing and the results of processing. The output device 35 is, for example, a display device (liquid crystal monitor, LCD (Liquid Crystal Display), graphic card, etc.) that visualizes the various information described above, a device that converts the various information described above into audio (audio output device (speaker, etc.)), or a device that converts the various information described above into text (printer, etc.). Note that, for example, the nesting response work support device 1 may be configured to input and output information to and from other devices via the communication device 36.

[0049] The input device 34 and the output device 35 constitute a user interface that receives and presents information.

[0050] The communication device 36 is a device that realizes communication with other devices. The communication device 36 is a wired or wireless communication interface that realizes communication with other devices via a communication network such as the Internet, and is, for example, a NIC (Network Interface Card), a wireless communication module, a USB module, or the like.

[0051] All or part of the nesting response work support device 1 may be realized using virtual information processing resources provided using virtualization technology, process space separation technology, or the like, such as a virtual server provided by a cloud system. All or part of the functions provided by the nesting response work support device 1 may be realized by a service provided by the cloud system via an API (Application Programming Interface), for example. The nesting response work support device 1 may also be configured as a system including multiple information processing devices connected to each other so that they can communicate with each other.

[0052] The above-mentioned functions of the nesting response work support device 1 are realized by the processor 31 of the nesting response work support device 1 reading and executing a program stored in the main memory device 32, or by the hardware (FPGA, ASIC, AI chip, etc.) that constitutes the nesting response work support device 1. The nesting response work support device 1 stores the above-mentioned various information (data), for example, as a database table or a file managed by a file system. Next, the processing performed by the nesting response work support device 1 will be described.

[0053] --Model learning process-- 6 is a flow diagram illustrating an example of model learning processing performed by the nesting response work support device 1. The model learning processing is started, for example, when a predetermined input is made to the nesting response work support device 1 by a worker or the like, or at a predetermined timing (a predetermined time, a predetermined time interval).

[0054] First, the learning unit 15 acquires information on the performance image and the corresponding correct label (s11). Specifically, the learning unit 15 acquires data of each record in the training data DB 200.

[0055] Then, the learning unit 15 performs preprocessing on the data acquired in s11 to generate feature quantities for a nest risk determination model (s13).

[0056] Specifically, for example, the learning unit 15 generates information about the time when the record image was captured (for example, information about the season) based on the capture date and time 202 in the training data DB 200.

[0057] Furthermore, for example, the learning unit 15 generates information on the area where the actual image was taken (for example, it may be an area based on an administrative district, or information on geographical characteristics that characterize the activity patterns of birds, such as residential areas or mountainous areas) based on the shooting location 204 in the teacher data DB 200 and predetermined map data.

[0058] Furthermore, the learning unit 15 may enlarge or reduce each actual image so that the same object in all actual images is composed of the same number of image elements.

[0059] Furthermore, the learning unit 15 may adjust each actual image by performing a projective transformation or the like to make the actual image into an image captured from a predetermined direction, or may correct the color tone or the like of each actual image.

[0060] The learning unit 15 performs machine learning on the record image acquired in s11, the feature amount calculated in s13, and the correct label, thereby creating a nest risk determination model (s15).

[0061] For example, the learning unit 15 inputs the image data 203 in the teacher data DB 200 acquired in s11 and the information on the season and region calculated in s13 into the nesting risk assessment model to acquire each output value and intermediate value corresponding to the actual image. When this process is executed for the first time, initial values are set in advance for the hyperparameters (described next) of the nesting risk assessment model.

[0062] The learning unit 15 optimizes the hyperparameters in the nest risk determination model so that each acquired output value approaches the nest area 205, power equipment area 206, power equipment type 207, and urgency level 208 registered in the teacher data DB 200 (s17). For example, the learning unit 15 adjusts hyperparameters such as weights between units (neurons) or coefficients in activation functions using a learning method such as backpropagation. This completes the model creation process.

[0063] The output value of the nesting risk assessment model may include information on the degree of urgency as well as information on its accuracy. In this case, the nesting risk assessment model can output one or more combinations of information on the degree of urgency and information on its accuracy.

[0064] The learning unit 15 may configure the nest risk assessment model using two trained models. For example, first, the learning unit 15 generates a first trained model that receives an actual image and outputs information on the nest area, the power equipment type, and the power equipment area in the actual image. That is, the learning unit 15 inputs the image data 203 in the training data DB 200 acquired in s11 into the first trained model, thereby acquiring each output value corresponding to the actual image. The learning unit 15 optimizes the hyperparameters in the nest risk assessment model so that each acquired output value approaches the nest area 205, the power equipment area 206, and the power equipment type 207 in the training data DB 200, respectively.

[0065] Second, the learning unit 15 generates a second trained model that receives the nest area, the type of power equipment, the area of the power equipment, the season, and the region, and outputs information on the level of urgency. That is, the learning unit 15 inputs the information on the nest area, the type of power equipment, and the area of the power equipment output by the first trained model into the second trained model, thereby acquiring each output value. The learning unit 15 optimizes the hyperparameters in the second trained model so that each acquired output value approaches the level of urgency 208 in the training data DB 200.

[0066] In this way, by configuring the nest risk assessment model with two trained models, it is possible to accurately recognize the areas of nests and each power facility and estimate the urgency, thereby enabling accurate determination of the response that should be taken in response to nesting and the timing of that response.

[0067] - Urgency determination process - 7 is a flow diagram illustrating an example of the urgency determination process. The urgency determination process is started, for example, when a predetermined input is made to the nest response work support device 1 by a worker or the like, or at a predetermined timing (a predetermined time, a predetermined time interval).

[0068] The image acquisition unit 11 acquires an image (site image) including the nest and utility pole for which the urgency level is to be determined (s31). For example, the image acquisition unit 11 reads an image taken by a camera or terminal held by a worker or the like via a recording medium or a communication network such as the Internet or a LAN (Local Area Network). The image acquisition unit 11 may also read an image taken by a surveillance camera installed on or near the utility pole.

[0069] The urgency determination unit 17 also receives input of information about the season and area in which the site image was taken from a worker or the like (s33). For example, the urgency determination unit 17 receives input of the season and area from a worker or the like while displaying the site image acquired in s31 on a screen. Note that the urgency determination unit 17 may generate information about the season and area in which the site image was taken from time information or location information included in the site image data by processing similar to that in s13.

[0070] The urgency determination unit 17 inputs the on-site image acquired in s31 and the season and area information input in s33 into the nest risk determination model, thereby acquiring urgency information output from the nest risk determination model (s35).The urgency determination unit 17 then displays the urgency information on an urgency determination screen (s37), which will be described next.This completes the urgency determination process.

[0071] <Urgency determination screen> 8 is an example of the urgency determination screen 500. The urgency determination screen 500 includes a display field 501 for an on-site image, a display field 503 for the season in which the on-site image was taken, a display field 505 for the region in which the on-site image was taken, and a display field 507 for the urgency determined by the nest risk determination model.

[0072] By referring to this urgency determination screen 500, the user can know what kind of action should be taken and when to take action against the nest formation captured in the on-site image.

[0073] As described above, the nest response work support device 1 of this embodiment receives an image (actual image) containing a power facility and a bird's nest, and outputs information on the response content and timing for the bird's nest by inputting the image (site image) into a trained model (nesting risk determination model), which outputs information (degree of urgency) on the response content and timing for the nest contained in the site image, so that workers who monitor or remove nests can know the specific response to the nest and the timing for that response simply by preparing an image of the site where the nest is being built.In other words, the nest response work support device 1 of this embodiment can provide information on the response to nests built in power facilities and the timing for that response.

[0074] Furthermore, the nesting response work support device 1 of this embodiment outputs information on the response content and response timing for nests contained in the on-site image by inputting information on the area and time when the on-site image was taken into a trained model (nesting risk determination model) to which information on the area and time when the image was taken is input. Because the nesting speed of birds can sometimes vary by region or season, by inputting the area and time when the on-site image was taken as feature quantities of the nesting risk determination model, workers and the like can accurately know the specific response that should be taken against nesting.

[0075] Furthermore, by inputting a site image into the nest risk assessment model, the nest response work support device 1 of this embodiment outputs information indicating that the nest included in the site image needs to be removed immediately, information indicating that the nest needs to be removed by a predetermined time in the future, or information indicating that the nest does not need to be removed. This allows workers and the like to know more specific measures to be taken against nests.

[0076] In addition, the nest response work support device 1 of this embodiment displays information on the response content and response timing for the nest contained in the site image on the screen, so workers and others can quickly understand the response that should be taken against the nest.

[0077] Furthermore, the nesting response work support device 1 of this embodiment outputs information on response and timing using a trained model (nesting risk determination model) consisting of a first trained model and a second trained model. That is, the nesting response work support device 1 of this embodiment inputs an image including a power facility and a bird's nest into the first trained model, which outputs information on the areas of the bird's nest and the power facility in the image. The first trained model then inputs the information on the areas of the bird's nest and the power facility in the image, outputting information on the response content and timing for the bird's nest. The second trained model then inputs the information on the areas of the bird's nest and the power facility output by the first trained model, outputting information on the response content and timing for the nest included in the site image. In this way, by using the second trained model with the information on the areas of the nest and the power facility recognized by the first trained model as input values, the response content and timing for the bird's nest can be more accurately estimated.

[0078] The above description of the embodiments is intended to facilitate understanding of the present invention, and is not intended to limit the present invention. The present invention may be modified or improved without departing from the spirit thereof, and the present invention includes equivalents thereof.

[0079] For example, the combination of each functional unit of the nest response business support device 1 described in this embodiment is just one example, and for example, some of the functional units may be provided in other functional units, or multiple functional units may be combined into one functional unit.

[0080] In addition, in this embodiment, it is assumed that the actual images and the on-site images are images of nests on utility poles, but they may also be images of nests on steel towers or other power facilities. Note that the facilities related to nesting may be used as one of the features of the nest risk assessment model.

[0081] In this embodiment, the urgency level, which is the output value of the nest risk assessment model, is set to the presence or absence of nest removal and the time until removal, but other types of response and timing may also be used as the urgency level. For example, the urgency level may be set to the strengthening of nest patrols and their duration, specific nest removal methods, etc.

[0082] Furthermore, the nest response work support device 1 may accept the specification of on-site images from multiple directions of the same utility pole and output urgency information for each on-site image. Because the configuration of power equipment installed on utility poles is complex, the images captured from different directions may differ significantly. Therefore, by outputting urgency information based on on-site images from multiple directions, workers can more accurately determine the urgency. [Explanation of symbols]

[0083] 1 Nest response support device, 11 Image acquisition unit, 13 Preprocessing unit, 15 Learning unit, 17 Urgency determination unit, 19 Information display unit

Claims

1. A memory for storing an image including a bird's nest and an electric power facility; an image acquisition process for acquiring the image; a processor that executes an urgency determination process that inputs an image including an area of an electric power facility where birds are using the electric power facility as a foothold, an area of the bird's nest located on the electric power facility, and an area of other electric power facility near the electric power facility, and outputs information about the action content and timing of the action to be taken with respect to the bird's nest by inputting the acquired image into a trained model that outputs information about the action content and timing of the action to be taken with respect to the bird's nest. A nest response support device equipped with the above.

2. The trained model is a model to which information on the area and time when the image was taken is further input, In the urgency determination process, the processor further inputs information about the area where the acquired image was taken and the time when the image was taken into the trained model, and outputs information about the response content and response timing for the nest included in the acquired image. The nesting response work support device according to claim 1.

3. The trained model is a model that outputs, as information on the response content and response timing for the bird's nest, any of information indicating that the nest needs to be immediately removed, information indicating that the nest needs to be removed by a predetermined time in the future, or information indicating that the nest does not need to be removed, In the urgency determination process, the processor inputs the acquired image into the trained model, and outputs, as information on the response content and response timing, any one of information indicating that the nest included in the acquired image needs to be immediately removed, information indicating that the nest needs to be removed by a predetermined time in the future, and information indicating that the nest does not need to be removed. The nesting response work support device according to claim 1.

4. The nest response work support device described in Claim 1, wherein the processor executes a display process that displays on a screen information regarding the response content and timing of response to the nest contained in the acquired image.

5. In the urgency determination process, the processor: an image including an area of an electric power facility where birds are using the facility as a foothold, an area of the bird's nest located on the electric power facility, and areas of other electric power facilities in the vicinity of the electric power facility is input, and information on the area of the bird's nest and the area of each piece of electric power facility in the image is output by inputting the acquired image into a first trained model; information on bird nest areas and areas of each piece of power equipment is input to a second trained model, which outputs information on the response content and response timing for the bird nests, and by inputting the output information on bird nest areas and areas of each piece of power equipment into the second trained model, information on the response content and response timing for the nests included in the acquired image is output. The nesting response work support device according to claim 1.

6. A method for supporting nest response operations using an information processing device including a memory for storing images including bird nests and power facilities, and a processor, the processor: an image acquisition process for acquiring the image; an urgency determination process in which an image including the area of the power equipment where the birds are using it as a foothold, the area of the bird's nest located on the power equipment, and the area of other power equipment in the vicinity of the power equipment is input, and information regarding the response content and timing of the response to the bird's nest is output by inputting the acquired image into a trained model; A method for supporting nest response operations.

Images