Sound analysis device, sound analysis method, and program

Abstract

A sound analysis device 1 according to one embodiment of the present invention comprises: an acquisition unit 131 that acquires information related to sounds acquired by a plurality of sound collection devices disposed in a predetermined region; and an output unit 133 that outputs information corresponding to the frequency of sound production by a living being in the region on the basis of the information acquired by the acquisition unit. The acquisition unit 131 acquires, for example, information indicating the arrival direction of the sound for each of the plurality of sound collection devices estimated on the basis of the sound that has arrived at each of the plurality of sound collection devices. The output unit 133 outputs the information corresponding to the frequency of sound production, for example, on the basis of a sound production position that is a position where specified sound production has been performed using the relationship between a plurality of arrival directions corresponding to the same sound that has arrived at the plurality of sound collection devices.

Description

Sound analysis device, sound analysis method, and program

[0001] The present invention relates to a sound analysis device, a sound analysis method, and a program for analyzing sound.

[0002] Patent document 1 discloses a device that is installed in a forest area, which determines whether or not a captured image contains a wild animal in response to a motion sensor detecting a moving object, and transmits data indicating that a wild animal has been detected in response to a determination that the image contains a wild animal.

[0003] Japanese Patent Application Laid-Open No. 2020-156400

[0004] In recent years, there has been a demand for measuring data on the ecology of organisms in natural environments in order to protect biodiversity. The technology described in Patent Document 1 performs detection using images captured by a camera, making it difficult to detect organisms at night or in the shadows of objects. Furthermore, because the camera's imaging range is limited, there is a problem in that it is costly to install a large number of cameras to detect organisms over a wide area of ​​the natural environment.

[0005] The present invention has been made in consideration of these points, and has as its object to reduce the cost of measuring data on living organisms in the natural environment.

[0006] A first aspect of the sound analysis device of the present invention includes an acquisition unit that acquires information related to sounds acquired by a plurality of sound collection devices arranged in a predetermined area, and an output unit that outputs information corresponding to the frequency of sounds made by living organisms in the area based on the information acquired by the acquisition unit.

[0007] The acquisition unit may acquire information indicating the direction of arrival of sound for each of the plurality of sound collection devices estimated based on the sound arriving at each of the plurality of sound collection devices, and the output unit may output information corresponding to the frequency of the pronunciation based on a pronunciation position, which is the position at which the pronunciation was made, identified using the relationship between multiple arrival directions corresponding to the same sound arriving at the plurality of sound collection devices.

[0008] The sound analysis device may further include an identification unit that identifies an index corresponding to the diversity of the organisms by summing the frequencies of the pronunciations of each of the multiple types of organisms, and the output unit may output the index as information corresponding to the frequencies of the pronunciations.

[0009] The identification unit may identify the index by weighting the frequency of the pronunciation of each of the plurality of types of the organism according to the rarity of the type.

[0010] The identification unit may identify the indicator by weighting the frequency of the sound for each of the plurality of types of the living thing in accordance with the danger of the type.

[0011] The identification unit may identify the index for each of a plurality of partial regions that make up the region by summing the frequencies of the pronunciations of each of a plurality of types of organisms in the partial region, and the output unit may output the index identified by the identification unit for each of the plurality of partial regions in association with the partial region.

[0012] The output unit may output information corresponding to the frequency of the pronunciation for each of a plurality of periods in association with the period.

[0013] The acquisition unit may acquire information indicating the state of each of the plurality of sound collection devices from the plurality of sound collection devices, and the output unit may output information indicating a route to reach one or more of the plurality of sound collection devices whose state satisfies a predetermined condition.

[0014] The output unit may output information indicating the presence of a specified creature in association with the route, on the condition that the acquisition unit acquires a sound emitted by the specified creature within a specified range including the route.

[0015] The sound analysis device may further include a generation unit that generates sound according to the information acquired by the acquisition unit, and the output unit may output the sound generated by the generation unit from a sound output unit possessed by one or more of the plurality of sound collection devices.

[0016] The generator may generate a sound corresponding to the type of the creature that made the sound.

[0017] The generation unit may generate a sound that conveys a specific intention to the type of organism that made the sound, and the output unit may output the sound generated by the generation unit from the sound output unit of one or more sound collection devices among the multiple sound collection devices to which the sound emitted by the type of organism has arrived.

[0018] A sound analysis method of a second aspect of the present invention includes the steps of: acquiring information relating to sounds picked up by a plurality of sound collection devices arranged in a predetermined area; and outputting information corresponding to the frequency of sounds made by living organisms in the area based on the information acquired in the acquiring step, executed by a processor.

[0019] A program of a third aspect of the present invention causes a processor to execute the steps of acquiring information relating to sounds picked up by a plurality of sound collection devices arranged in a predetermined area, and outputting information corresponding to the frequency of sounds made by living organisms in the area based on the information acquired in the acquiring step.

[0020] The present invention has the effect of reducing the cost of measuring data on living organisms in the natural environment.

[0021] 1 is a schematic diagram of a sound analysis system according to an embodiment. FIG. 2 is a block diagram of a sound analysis system according to an embodiment. FIG. 3 is a schematic diagram for explaining a plurality of sound collection devices arranged in an area. FIG. 4 is a schematic diagram for explaining a method for identifying the direction from which a sound comes. FIG. 5 is a schematic diagram for explaining a method for identifying a sound production position. FIG. 6 is a schematic diagram for explaining a method for identifying a production frequency. FIG. 7 is a schematic diagram of an information terminal displaying information including a diversity index. FIG. 8 is a schematic diagram for explaining a method for outputting information indicating whether or not an abnormality has occurred in a sound collection device or an environment including the sound collection device. FIG. 9 is a schematic diagram for explaining a method for outputting generated sound from a sound collection device. FIG. 10 is a diagram showing a flowchart of an exemplary sound analysis method executed by a sound analysis device according to an embodiment. FIG. 11 is a schematic diagram for explaining a method for determining biodiversity using the movement paths of organisms.

[0022] 1 is a schematic diagram of a sound analysis system S according to this embodiment. The sound analysis system S includes a sound analysis device 1, a sound collection device 2, and an information terminal 3. The sound analysis system S may also include other devices such as a server and a terminal.

[0023] The sound analysis device 1 is a computer that analyzes sounds emitted by living organisms in a predetermined region R and provides the analysis results to a user. The region R is an area to be analyzed, such as an area in a natural environment where multiple living organisms live. In this embodiment, the region R is an area on land, but it may also be an area in the air or underwater.

[0024] A living thing is a living organism that resides in area R and makes sounds within area R. A living thing makes sounds using, for example, a function of the organism's own body (such as vocal organs). A living thing may also make sounds using an object other than the organism (such as the sound of hitting an object).

[0025] The sound analysis device 1 analyzes the sound acquired by the sound collection device 2 and outputs the analysis results to the information terminal 3. The sound analysis device 1 is connected to the sound collection device 2 and the information terminal 3 via a network such as a local area network or the Internet.

[0026] The sound collection device 2 is a device that acquires sounds emitted by living organisms. Multiple sound collection devices 2 are arranged in the region R. Each of the multiple sound collection devices 2, for example, has a microphone array including a sound collection unit such as multiple microphones arranged in different orientations. The microphone array includes, for example, multiple microphones (e.g., eight) arranged at equal intervals on the same circumference in a horizontal plane relative to the ground. The sound analysis device 1 identifies the position from which the sound is emitted by estimating the direction from which the sound arrives at each of the multiple sound collection devices 2 based on the sound collected using the microphone array. The sound collection device 2 transmits the sound acquired using the microphone array to the sound analysis device 1 as sound data.

[0027] The sound collection device 2 may also have a sound output unit such as a speaker for outputting sound, and a measurement unit for measuring data indicating the state of the sound collection device 2. The measurement unit includes, for example, an acceleration sensor, a temperature sensor, etc.

[0028] The information terminal 3 is a computer used by a user. The user is a person who obtains the analysis results from the sound analysis device 1. The information terminal 3 is, for example, a smartphone, a tablet terminal, a personal computer, etc. The information terminal 3 has a display unit such as a liquid crystal display. The information terminal 3 displays the information received from the sound analysis device 1 on the display unit.

[0029] An overview of the process of analyzing sound by the sound analysis system S according to this embodiment will be described below. Each of the multiple sound collection devices 2 acquires sound emitted in the region R and transmits sound information related to the acquired sound to the sound analysis device 1. The sound information includes, for example, at least one of data generated by encoding the sound and data indicating the direction from which the sound is coming. The sound analysis device 1 acquires the sound information transmitted by each of the multiple sound collection devices 2 arranged in the region R.

[0030] The sound analysis device 1 identifies the frequency of sounds made by living organisms in the region R based on the acquired sound information, and outputs information corresponding to the identified frequency of sounds to the information terminal 3. The information corresponding to the frequency of sounds includes, for example, a diversity index corresponding to the diversity of living organisms in the region R, calculated using the frequency of sounds.

[0031] In this way, the sound analysis system S outputs information corresponding to the frequency of sounds made by living organisms in the area R, based on information relating to sounds acquired by the multiple sound collection devices 2 arranged in the area R. This allows the sound analysis system S to provide data on the distribution of living organisms without installing a large number of cameras in the natural environment, thereby reducing the cost of measuring data on living organisms in the natural environment.

[0032] In other words, when capturing images with a camera, the range that can be captured is limited by the angle of view, and if there is an obstruction, it is not possible to capture what is beyond that, so the range that can be observed by a single camera is limited. Therefore, in order to provide data on the distribution of organisms, it is necessary to install many cameras.

[0033] In contrast, when sound is acquired using the sound collection device 2, sounds from all directions around the area can be acquired, and sounds emitted from beyond obstructions can be acquired, so the range that can be observed with one sound collection device 2 is wider than when a camera is used. Therefore, by using the sound collection devices 2, the sound analysis system S can provide data on the distribution of organisms with fewer devices than when cameras are used, and can reduce the cost of measuring data on organisms in natural environments.

[0034] Furthermore, while cameras have difficulty capturing images of organisms that are active at night or that hide in the shadows, the sound analysis system S can easily measure data on organisms that are active at night or that hide in the shadows by using the sounds that the organisms make.

[0035] [Configuration of Sound Analysis System S] Figure 2 is a block diagram of the sound analysis system S according to this embodiment. In Figure 2, arrows indicate main data flows, and data flows other than those shown in Figure 2 may exist. In Figure 2, each block indicates a functional configuration rather than a hardware (device) configuration. Therefore, the blocks shown in Figure 2 may be implemented in a single device, or may be implemented separately in multiple devices. Data may be exchanged between blocks via any means, such as a data bus, a network, or a portable storage medium.

[0036] The sound analysis device 1 includes a communication unit 11, a storage unit 12, and a control unit 13. The sound analysis device 1 may be configured by connecting two or more physically separate devices via a wired or wireless connection. The sound analysis device 1 may also be configured by a cloud, which is a collection of computer resources.

[0037] The communication unit 11 is a communication interface for transmitting and receiving data between the sound collection device 2 and the information terminal 3 via the network. The communication unit 11 notifies the control unit 13 of data received from the sound collection device 2 and the information terminal 3 via the network. The communication unit 11 also transmits data output from the control unit 13 to the sound collection device 2 and the information terminal 3 via the network.

[0038] The storage unit 12 is a computer-readable non-transitory storage medium including a ROM (Read Only Memory), a RAM (Random Access Memory), a hard disk drive, etc. The storage unit 12 stores in advance programs to be executed by the control unit 13. The storage unit 12 may be provided outside the sound analysis device 1, in which case data may be exchanged between the storage unit 12 and the control unit 13 via a network.

[0039] The control unit 13 has an acquisition unit 131, an identification unit 132, an output unit 133, and a generation unit 134. The control unit 13 is a processor such as a CPU (Central Processing Unit), and functions as the acquisition unit 131, the identification unit 132, the output unit 133, and the generation unit 134 by executing a program stored in the storage unit 12. Furthermore, at least some of the functions of the control unit 13 may be realized by the control unit 13 executing a program executed via a network.

[0040] The processing executed by the sound analysis system S will be described in detail below. Fig. 3 is a schematic diagram for explaining a plurality of sound collection devices 2 arranged in an area R. The plurality of sound collection devices 2 are arranged in advance within the area R. The storage unit 12 of the sound analysis device 1 stores information indicating the position (e.g., coordinates) of each of the plurality of sound collection devices 2.

[0041] The sound collection device 2 acquires (collects) sounds that arrive at the sound collection device 2. The sound collection device 2 generates sound information related to the acquired sounds and transmits the generated sound information to the sound analysis device 1. The sound collection device 2 transmits, as sound information, data generated by, for example, performing a known encoding process on sound data indicating sounds collected using a microphone array. The sound collection device 2 constantly transmits sound information to the sound analysis device 1, or transmits sound information for a predetermined period (one hour, one day, etc.) collectively to the sound analysis device 1.

[0042] Furthermore, the sound collection device 2 may generate measurement information regarding the sound collection device 2 and transmit the generated measurement information to the sound analysis device 1. The sound collection device 2 transmits, for example, data such as acceleration or temperature measured using a measurement unit as measurement information.

[0043] The sound collection device 2 transmits sound information and measurement information to the sound analysis device 1 using wireless communication such as LTE (Long Term Evolution) or 5G (5th Generation). Alternatively, the sound collection device 2 may transmit sound information and measurement information to a mobile communication device (vehicle, drone, ship, etc.) using wireless communication with a relatively narrow communication range such as a wireless LAN (Local Area Network), and after the communication device has moved, the communication device may transmit the sound information and measurement information to the sound analysis device 1 using wireless communication with a relatively wide communication range such as LTE or 5G. This allows the sound collection device 2 to transmit sound information and measurement information to the sound analysis device 1 via the mobile communication device, even if the sound collection device 2 is located in an area (such as deep in the mountains) where communication is not possible using LTE, 5G, etc.

[0044] In the sound analysis device 1, the acquisition unit 131 acquires sound information transmitted by each of the multiple sound collection devices 2. The acquisition unit 131 also acquires measurement information transmitted by each of the multiple sound collection devices 2. The acquisition unit 131 stores the acquired sound information and measurement information in the storage unit 12 in association with identification information for identifying the sound collection device 2 that sent the information.

[0045] The acquisition unit 131 identifies the direction from which the sound arrives for each of the plurality of sound collection devices 2 based on the sound that has arrived at each of the plurality of sound collection devices 2 indicated by the acquired sound information.

[0046] 4 is a schematic diagram illustrating a method for identifying the direction of sound arrival. The acquisition unit 131, for example, at predetermined intervals (e.g., every 10 milliseconds to 1000 milliseconds), identifies the direction of sound arrival D of each of the multiple sound collection devices 2 indicated by the acquired sound information. The acquisition unit 131 performs a known sound source localization process on multiple channels of sound collected by a microphone array provided in the sound collection device 2. The sound source localization process is a process for estimating the position of a sound source contained in the sound acquired by the acquisition unit 131. Through the sound source localization process, the acquisition unit 131 acquires a confidence distribution P1 indicating a distribution of confidence that a sound source exists based on the position of the sound collection device 2. The confidence is a value corresponding to the likelihood that a sound source exists at that position, and may be, for example, a probability.

[0047] 4 shows a reliability distribution P1 acquired by the acquisition unit 131 based on sounds arriving at each of the three sound collection devices 2. The vertical and horizontal axes of the reliability distribution P1 correspond to coordinates within the region R. The reliability distribution P1 indicates that the brighter the color of each position (coordinate) (the closer to white), the higher the reliability that a sound source exists, and that the darker the color of each position (the closer to black), the lower the reliability that a sound source exists.

[0048] Since the microphone array cannot determine the distance from the sound collector 2 to the sound source, regions of the same reliability are distributed linearly (radially) with the sound collector 2 as the reference in the reliability distribution P1. Furthermore, since the sound attenuates more as the distance between the sound collector 2 and the sound source increases, the reliability distribution P1 may be weighted so that the value decreases the farther away from the sound collector 2 and increases the closer to the sound collector 2. Since the reliability that a sound source exists increases on the line connecting the sound collector 2 and the sound source, a linear region in the reliability distribution P1 where the reliability is equal to or greater than a predetermined value indicates the direction D from which the sound arrives relative to the sound collector 2. The direction of arrival D is not limited to a straight line including the position of the sound collector 2, but may be expressed as a region having a width of a predetermined angle or length based on the position of the sound collector 2.

[0049] In this embodiment, the sound analysis device 1 identifies the arrival direction, but the arrival direction may also be identified based on sounds acquired by each of the multiple sound collection devices 2 using a microphone array. In this case, in the sound analysis device 1, the acquisition unit 131 receives, from each of the multiple sound collection devices 2, sound information indicating the arrival direction identified by the sound collection device 2.

[0050] The identification unit 132 identifies the sound generation position, which is the position where sound generation was performed within the region R, for each predetermined period (for example, every 10 to 1000 milliseconds) using the relationship between multiple arrival directions corresponding to the same sound that arrived at multiple sound collection devices 2. The same sound is a sound emitted from one sound source at one point in time.

[0051] Fig. 5 is a schematic diagram for explaining a method for identifying a sound production position. The identification unit 132 overlaps multiple reliability distributions P1 generated from sounds arriving at multiple sound collection devices 2. The identification unit 132 overlaps the multiple reliability distributions P1 by, for example, calculating the sum or product of the reliability indicated by the multiple reliability distributions P1 at each position within the region R. Fig. 5 shows a reliability distribution P2 generated by overlapping the three reliability distributions P1 illustrated in Fig. 4.

[0052] The identification unit 132 identifies the sound production position using a reliability distribution P2 obtained by superimposing a plurality of reliability distributions P1. The sound production position may be represented by a single point within the region R, or may be represented by a region having an area within the region R. For example, the identification unit 132 identifies, as the sound production position, a position or region in the reliability distribution P2 whose reliability is equal to or greater than a predetermined value.

[0053] The position where the multiple arrival directions D indicated by the multiple reliability distributions P1 intersect is a position with high reliability in the reliability distribution P2 obtained by superimposing the multiple reliability distributions P1. In the example of Fig. 5, the identification unit 132 identifies the position D1 where the multiple straight lines along the multiple arrival directions D intersect as the sound production position. Furthermore, when the arrival direction D is a region with width, the identification unit 132 may identify the region where the multiple regions extending along the multiple arrival directions D intersect as the sound production position.

[0054] In this way, the sound analysis device 1 identifies the sound production position based on the direction D of sound arrival relative to multiple sound collection devices 2, and therefore can identify the sound production position with high accuracy even when the distance from one sound collection device 2 to the sound source cannot be identified.

[0055] The identification unit 132 may identify the sound production position using sounds that arrive at multiple sound collection devices 2 equipped with a single microphone, instead of sounds that arrive at multiple sound collection devices 2 equipped with a microphone array. In this case, multiple sound collection devices 2 are arranged at predetermined intervals in the region R. When a living creature makes a sound within the region R, each sound collection device 2 picks up a sound with higher intensity the closer it is to the living creature, and picks up a sound with lower intensity the farther it is from the living creature.

[0056] The identification unit 132 compares the intensities of sounds acquired simultaneously by the multiple sound collection devices 2, and identifies the sound production position based on the sound collection device 2 with the highest acquired sound intensity or the positions of multiple sound collection devices 2 with acquired sound intensities equal to or greater than a threshold. This allows the sound analysis device 1 to identify the sound production position even when using sound collection devices 2 that do not have microphone arrays.

[0057] The identification unit 132 estimates the type of organism that made the sound at the identified sound production position. The identification unit 132 acquires, for example, a machine learning model pre-stored in the storage unit 12 and configured to output the type of organism that produced the input sound. The machine learning model is generated, for example, by machine learning training data indicating the sounds produced by each of multiple types of organisms and the types using a known machine learning process, and is stored in the storage unit 12. Because the sounds produced by organisms are also affected by the time of day and the region, machine learning may be performed to generate the machine learning model using training data that further includes information such as the time the organism produced the sound, the region where the sound collection device is located, and the climate of the region. The identification unit 132 inputs the sound used to identify the sound production position into the machine learning model, and estimates the type of organism that made the sound at the sound production position based on the type of organism output by the machine learning model.

[0058] The identification unit 132 uses the identified sound production positions and the estimated species of organisms to identify the frequency of sound production by each of the multiple species of organisms in each of the multiple partial regions R1 that make up the region R. The multiple partial regions R1 are, for example, multiple regions generated by dividing the region R into regions of a predetermined size and shape. It is desirable that the multiple partial regions R1 do not overlap each other.

[0059] 6 is a schematic diagram illustrating a method for identifying the frequency of sounds produced by a living creature at one or more sound production positions included in each of a plurality of partial regions R1. For example, the identification unit 132 counts the number of sounds produced by the living creature at the one or more sound production positions for each type of living creature.

[0060] The identification unit 132 calculates the number of pronunciations per unit time (e.g., one day, one month, etc.) by dividing the total number of pronunciations for each combination of each of the multiple partial regions R1 and each of the multiple types of organisms by the period of time being counted. The identification unit 132 identifies the number of pronunciations per unit time calculated for each combination of each of the multiple partial regions R1 and each of the multiple types of organisms as the frequency of pronunciation for that type of organism at that sound generation position.

[0061] 6, the identification unit 132 identifies that a boar, a deer, and a sparrow have made sounds with their respective sound frequencies in a certain partial region R1. The identification unit 132 is not limited to the specific method shown here, and may identify the sound frequencies of each of a plurality of types of living creatures in each of a plurality of partial regions R1 using other methods.

[0062] The identification unit 132 uses the pronunciation frequency of each of the multiple types of organisms identified for each of the multiple partial regions R1 to identify a diversity index corresponding to the diversity of organisms in the partial region R1. The identification unit 132 identifies the diversity index, for example, by assigning a predetermined weight to the pronunciation frequency of each of the multiple types of organisms and then summing the weighted values.

[0063] The diversity index has a larger value as the diversity in the partial region R1 is greater (for example, the number and types of organisms is greater), and a smaller value as the diversity is smaller (for example, the number and types of organisms is fewer).Furthermore, the diversity index may have a smaller value as the diversity in the partial region R1 is greater, and a larger value as the diversity is smaller.

[0064] The specifying unit 132 calculates a diversity index I for each of the plurality of partial regions R1 using, for example, the following formula (1).

[0065]

[0066] In formula (1), sn (n is a natural number within a predetermined range) indicates each of the multiple organism types, wsn indicates a weight, and fsn indicates a pronunciation frequency. The pronunciation frequency indicated by fsn is the pronunciation frequency identified by the identification unit 132 for the organism type sn.

[0067] The type of organism represented by sn may be a "species" on a phylogenetic tree hierarchy, or may be other classifications such as a "genus" or a "family." The types of organisms are defined, for example, such that s1 represents a boar, s2 represents a deer, and s3 represents a sparrow. The types of organisms may also be other groups that can be identified by the patterns of sounds they emit.

[0068] The weight indicated by wsn is a value by which the pronunciation frequency fsn of the organism type sn is multiplied, and is stored in advance in the storage unit 12. For example, the weight is a value greater than 0 when the organism type corresponding to the pronunciation frequency is the target of aggregation (for example, pronunciation by an organism that may inhabit the area R), and is 0 when it is not the target of aggregation (for example, pronunciation by a human). This allows the sound analysis system S to reflect the pronunciation frequency in the diversity index only for the organism type that is the target of aggregation.

[0069] The weight may be 0, for example, if the frequency of the sound is likely to be identified from sounds not produced by living organisms (e.g., noise such as wind noise or automobile noise). Whether a sound is produced by a living organism is determined, for example, by a machine learning model that estimates the type of organism. This allows the sound analysis system S to identify a diversity index excluding sounds not produced by living organisms.

[0070] The weight may be a value according to the rarity of the species of organism corresponding to the frequency of the sound or its impact on the ecosystem. In this case, for example, the weight is greater the rarity of the species of organism (more rare), and is smaller the rarity of the species of organism (less rare). Furthermore, the weight may be smaller the rarity of the species of organism (more rare), and is larger the rarity of the species of organism (less rare). Furthermore, the weight may be greater or smaller for keystone species (species known to be important in the ecosystem) that have a large impact on the ecosystem, and smaller or larger for other species. This allows the sound analysis system S to reflect the presence and quantity of organisms that contribute to biodiversity in the biodiversity index.

[0071] The weight may be a value according to the danger of the type of organism corresponding to the frequency of the sound. In this case, for example, the weight is a larger value the more dangerous the type of organism (more dangerous), and a smaller value the less dangerous the type of organism (less dangerous). Furthermore, the weight may be a smaller value the more dangerous the type of organism (more dangerous), and a larger value the less dangerous the type of organism. This allows the sound analysis system S to reflect the presence or absence and quantity of dangerous organisms that may damage diversity, such as pests or specified invasive species, in the diversity index.

[0072] The weight may be a value according to the value as a tourist resource of the type of organism corresponding to the frequency of the pronunciation. In this case, for example, the weight is greater the more valuable (more popular) the organism as a tourist resource, and is smaller the less valuable (less popular) the organism as a tourist resource. Also, for example, the weight may be smaller the more valuable the organism as a tourist resource, and is larger the less valuable the organism as a tourist resource. This allows the sound analysis system S to reflect the presence or absence and quantity of organisms that have tourist resource value in the diversity index.

[0073] In this embodiment, the specifying unit 132 specifies a diversity index for each of the plurality of partial regions R1, but may specify a diversity index for the entire region R in the same manner.

[0074] The determination unit 132 may determine a diversity index corresponding to the frequency of pronunciation for each of a plurality of periods. The plurality of periods may be, for example, seasons such as the rainy season and the dry season, or a period of a predetermined length such as one month. In this case, the determination unit 132 determines the frequency of pronunciation using sounds that arrive at the sound collection device 2 for each of the plurality of periods, and determines a diversity index for each of the periods using the determined frequency of pronunciation.

[0075] The output unit 133 outputs information corresponding to the pronunciation frequency identified by the identification unit 132. The output unit 133 transmits, for example, display information for displaying information including a diversity index for each of the plurality of partial regions R1 identified by the identification unit 132 as information corresponding to the pronunciation frequency to the information terminal 3. The information terminal 3 displays the information including the diversity index for each of the plurality of partial regions R1 on the display unit in accordance with the display information transmitted by the sound analysis device 1.

[0076] 7 is a schematic diagram of an information terminal 3 displaying information including a diversity index. The output unit 133 outputs the diversity index identified by the identification unit 132 for each of multiple partial regions R1 to the information terminal 3 in association with the partial region R1. In the example of FIG. 7 , the output unit 133 displays a heat map on the information terminal 3, generated by assigning a color or pattern to each of the multiple partial regions R1 in the map of region R according to the diversity index of the partial region R1. This allows the sound analysis system S to enable the user to easily grasp the biological diversity of each partial region R1 that constitutes region R.

[0077] Furthermore, when the diversity index identified by the identification unit 132 for each of the multiple partial regions R1 satisfies a predetermined condition indicating that the diversity index is abnormal (such as low diversity), the output unit 133 displays alert information A indicating that the diversity index is abnormal on the information terminal 3. The condition indicating that the diversity index is abnormal includes, for example, the diversity index being equal to or greater than a predetermined reference value or being equal to or less than a predetermined reference value. Furthermore, the output unit 133 may determine whether or not a statistical value (such as an average value or a median value) of the diversity indexes of the multiple partial regions R1 that constitute the region R satisfies a predetermined condition. This allows the sound analysis system S to easily determine whether or not the diversity index of the entire region R or each partial region R1 is normal.

[0078] When the identification unit 132 identifies information (diversity index) corresponding to the frequency of pronunciation for each of a plurality of periods (rainy season, dry season, etc.), the output unit 133 may output the diversity index for each of the plurality of periods in association with the period. This enables the sound analysis system S to easily grasp changes in the diversity index over a plurality of periods.

[0079] The output unit 133 may output information indicating whether or not an abnormality has occurred in the sound collection device 2 or the environment including the sound collection device 2, based on the measurement information of the sound collection device 2 acquired by the acquisition unit 131. Fig. 8 is a schematic diagram for explaining a method of outputting information indicating whether or not an abnormality has occurred in the sound collection device 2 or the environment including the sound collection device 2.

[0080] The identification unit 132 determines, for example, whether or not the state (acceleration, temperature, etc.) indicated by the measurement information acquired from each of the multiple sound collection devices 2 satisfies a predetermined condition indicating the occurrence of an abnormality. The condition indicating the occurrence of an abnormality is, for example, that the acceleration or temperature indicated by the measurement information is equal to or greater than a predetermined reference value.

[0081] When the identification unit 132 determines that the state of any one of the multiple sound collection devices 2 satisfies the condition, the output unit 133 causes the information terminal 3 to display alert information A indicating that an abnormality has occurred in the sound collection device 2 or the environment including the sound collection device 2. In this way, the sound analysis system S can notify the user that an abnormality has occurred in the sound collection device 2 or the environment including the sound collection device 2, based on the data measured by each of the multiple sound collection devices 2.

[0082] If the identification unit 132 determines that the state of any one of the multiple sound collection devices 2 satisfies a condition indicating that an abnormality has occurred, the output unit 133 may output information indicating a route to reach one or more sound collection devices 2 among the multiple sound collection devices 2 that satisfy the condition.

[0083] In this case, the identification unit 132 executes, for example, a known route search process to identify a route for arriving at one or more sound collection devices 2 whose states satisfy a condition after departing from the departure point. The departure point is, for example, the current location of the information terminal 3 or a base associated with the multiple sound collection devices 2.

[0084] The output unit 133 transmits display information for displaying the route identified by the identification unit 132 to the information terminal 3. The information terminal 3 displays, on the display unit, information including routes for sequentially reaching one or more sound collection devices 2 whose states satisfy the conditions, in accordance with the display information transmitted by the sound analysis device 1.

[0085] 8 , the output unit 133 causes the information terminal 3 to display a map of an area R on which the starting point, the positions of the plurality of sound collection devices 2, and the route P are superimposed. This enables the sound analysis system S to enable the user to easily reach a sound collection device 2 that has been determined to have an abnormality based on data measured by each of the plurality of sound collection devices 2.

[0086] In addition, the output unit 133 may output information indicating the presence of a specified organism in association with the path identified by the identification unit 132, on the condition that the acquisition unit 131 has acquired a sound emitted by a specified organism around the path identified by the identification unit 132.

[0087] In this case, the identification unit 132 extracts the type of organism that made the sound at the sound generation position included within a predetermined range of the identified route (e.g., within 100 meters from the route) from, for example, the combination of the sound generation position and the type of organism. The identification unit 132 determines whether or not a predetermined organism type (e.g., a dangerous organism such as a bear) is included in one or more of the extracted organism types. When the identification unit 132 determines that a predetermined organism type is included in one or more of the extracted organism types, the output unit 133 displays, on the information terminal 3, alert information A indicating the presence of the predetermined organism along with the route P. In this way, the sound analysis system S can notify the user that a predetermined organism such as a bear may be present around the route to reach the sound collection device 2, based on the sounds that arrive at each of the multiple sound collection devices 2.

[0088] The output unit 133 may output the generated sound generated based on the sound information acquired by the acquisition unit 131 from the sound collection device 2. Figure 9 is a schematic diagram for explaining a method of outputting the generated sound from the sound collection device 2.

[0089] The generation unit 134 generates a generated sound corresponding to, for example, the type of living organism that emitted the sound that arrived at the sound collection device 2 indicated by the sound information acquired by the acquisition unit 131. The generated sound is, for example, a sound that conveys a specific intention to a specific type of living organism. The sound that conveys a specific intention is, for example, a sound that attracts living organisms, a sound that repels living organisms, or a sound that causes living organisms to take some other action.

[0090] The generation unit 134 acquires, for example, a machine learning model that is stored in advance in the storage unit 12 and configured to output a sound that corresponds to the input type of organism. The machine learning model is generated, for example, by using a known machine learning process to learn training data that indicates multiple types of organisms and sounds emitted by those types of organisms (such as attracting sounds or repelling sounds), and is stored in the storage unit 12. The generation unit 134 inputs the type of organism estimated by the identification unit 132 into the machine learning model, and generates the sound output by the machine learning model as a generated sound that corresponds to the type of organism.

[0091] The output unit 133 outputs the generated sound generated by the generation unit 134 from a sound output unit included in one or more of the sound collection devices 2. The output unit 133, for example, transmits generated sound information for outputting the generated sound to the sound collection device 2. The sound collection device 2 outputs the generated sound using the sound output unit in accordance with the generated sound information transmitted by the sound analysis device 1.

[0092] Here, it is desirable that the output unit 133 output the generated sound from a sound output unit of one or more sound collection devices 2 from which a sound emitted by a type of organism corresponding to the generated sound has arrived, among the multiple sound collection devices 2. In this case, the output unit 133 selects one or more sound collection devices 2 from which a sound emitted by a type of organism corresponding to the generated sound has arrived, based on the type of organism estimated by the identification unit 132, and transmits the generated sound information to the selected one or more sound collection devices 2. In this way, the sound analysis system S can output a generated sound corresponding to the type of organism that may be present near the sound collection device 2 from the sound collection device 2, and thereby intervene in the behavior of that type of organism.

[0093] For example, on the condition that the identification unit 132 determines that an abnormality has occurred in any of the sound collection devices 2 or in the environment including the sound collection devices 2, the output unit 133 may cause the sound output unit of the sound collection device 2 to output a generated sound that is a sound that repels living organisms of a type corresponding to the sound that has arrived at the sound collection device 2. In this way, when some abnormality (impact, fire, etc.) occurs in the sound collection device 2 or the environment including the sound collection device 2, the sound analysis system S can output a sound to repel living organisms of a type that may be present around the sound collection device 2, thereby preventing danger to the living organisms.

[0094] 10 is a diagram showing a flowchart of an exemplary sound analysis method executed by the sound analysis device 1 according to this embodiment. The acquisition unit 131 acquires sound information and measurement information transmitted by each of the multiple sound collection devices 2, and stores the acquired sound information and measurement information in the storage unit 12 in association with identification information for identifying the sound collection device 2 that transmitted the information (S11).

[0095] The acquisition unit 131 identifies the direction from which the sound arrives for each of the plurality of sound collection devices 2, based on the sound that has arrived at each of the plurality of sound collection devices 2 indicated by the acquired sound information. The identification unit 132 identifies the sound production position, which is the position at which sound production was made within the region R, for each predetermined period of time, using the relationship between the plurality of arrival directions corresponding to the same sound that has arrived at the plurality of sound collection devices 2 (S12).

[0096] The identification unit 132 estimates the type of organism that made the sound at the identified pronunciation position (S13). The identification unit 132 acquires, for example, a machine learning model that is pre-stored in the storage unit 12 and is configured to output the type of organism that made the input sound. The identification unit 132 inputs the sound used to identify the pronunciation position into the machine learning model, and estimates the type of organism that was output by the machine learning model as the type of organism that made the sound at the pronunciation position.

[0097] The identification unit 132 uses the identified sound production location and the estimated type of organism to identify the frequency of sound production by each of the multiple types of organisms in each of the multiple partial regions R1 that make up the region R (S14).

[0098] The identification unit 132 identifies a diversity index corresponding to the diversity of organisms in each partial region R1 by using the pronunciation frequency of each of the multiple types of organisms identified for each partial region R1 (S15). The identification unit 132 identifies the diversity index by, for example, adding up the pronunciation frequencies of each of the multiple types of organisms after assigning a predetermined weight to them.

[0099] The output unit 133 outputs information corresponding to the pronunciation frequency identified by the identification unit 132 (S16). The output unit 133 transmits, for example, display information for displaying information including a diversity index for each of the plurality of partial regions R1 identified by the identification unit 132 as information corresponding to the pronunciation frequency to the information terminal 3. The information terminal 3 displays the information including the diversity index for each of the plurality of partial regions R1 on the display unit in accordance with the display information transmitted by the sound analysis device 1.

[0100] Furthermore, the output unit 133 may output information indicating whether or not an abnormality has occurred in the sound collection device 2 or the environment including the sound collection device 2, based on the measurement information of the sound collection device 2 acquired by the acquisition unit 131. Furthermore, when the identification unit 132 determines that the state of any one of the multiple sound collection devices 2 satisfies a condition indicating that an abnormality has occurred, the output unit 133 may output information indicating a route to reach one or more sound collection devices 2 among the multiple sound collection devices 2 that satisfy the condition.

[0101] Effect of this embodiment The sound analysis system S according to this embodiment outputs information corresponding to the frequency of sounds made by living organisms in the area R, based on information relating to sounds acquired by a plurality of sound collection devices 2 arranged in the area R. This allows the sound analysis system S to provide data on the distribution of living organisms without installing a large number of cameras in the natural environment, thereby reducing the cost of measuring data on living organisms in the natural environment. Furthermore, by using sounds emitted by living organisms, the sound analysis system S can easily measure data on living organisms that are active at night or hiding in the shadows.

[0102] [Modification] In the above-described embodiment, the sound analysis system S determines the diversity of living organisms using a diversity index, whereas in this modification, the diversity of living organisms is determined using the migration paths of living organisms. Below, differences from the above-described embodiment will be mainly described.

[0103] 11 is a schematic diagram illustrating a method for determining biodiversity using the movement paths of organisms. The identification unit 132 identifies the individual organism that made the sound in addition to the sound production position. The identification unit 132 identifies the individual organism that made the sound by, for example, performing a known individual identification process on the sound information acquired by the acquisition unit 131, which identifies the individual organism using features (waveform, frequency, etc.) of the audio data.

[0104] The identification unit 132 identifies the movement path M of each of the identified individuals of the multiple organisms by connecting the multiple sound production positions of each of the identified individuals in chronological order. The identification unit 132 identifies an area of ​​a predetermined size that includes a point at which the movement paths M of different individuals equal to or greater than a predetermined reference value (e.g., 2 or more) intersect as a high diversity area N. The identification unit 132 may also identify an area of ​​a predetermined size that includes a point at which the movement paths M of different types of organisms equal to or greater than a predetermined reference value (e.g., 2 or more) intersect as a high diversity area N.

[0105] According to this modification, the sound analysis system S tracks the movement paths of living organisms based on sound-related information acquired by multiple sound collection devices 2 arranged in the area R, and identifies areas of high diversity where multiple individual organisms or multiple types of organisms pass through. This allows the sound analysis system S to provide data on the distribution of organisms without installing a large number of cameras in the natural environment, thereby reducing the cost of measuring data on organisms in the natural environment.

[0106] The present invention has been described above using embodiments, but the technical scope of the present invention is not limited to the scope described in the above embodiments, and various modifications and changes are possible within the scope of the gist of the present invention. For example, all or part of the device can be configured by functionally or physically distributing or integrating in any unit. Furthermore, new embodiments resulting from any combination of multiple embodiments are also included in the embodiments of the present invention. The effects of the new embodiments resulting from the combination also have the effects of the original embodiments.

[0107] The processor of the sound analysis device 1 executes each step (process) included in the sound analysis method shown in Fig. 10. That is, the processor of the sound analysis device 1 executes a program for executing the sound analysis method shown in Fig. 10, thereby executing the sound analysis method shown in Fig. 10. Some of the steps included in the sound analysis method shown in Fig. 10 may be omitted, the order of the steps may be changed, or multiple steps may be performed in parallel.

[0108] S Sound analysis system 1 Sound analysis device 11 Communication unit 12 Storage unit 13 Control unit 131 Acquisition unit 132 Identification unit 133 Output unit 134 Generation unit 2 Sound collection device 3 Information terminal

Claims

1. A sound analysis device having: an acquisition unit that acquires information related to sounds acquired by a plurality of sound collection devices arranged in a predetermined area; and an output unit that outputs information corresponding to the frequency of sounds made by living organisms in the area based on the information acquired by the acquisition unit.

2. The sound analysis device described in claim 1, wherein the acquisition unit acquires information indicating the direction of arrival of sound for each of the multiple sound collection devices estimated based on the sound arriving at each of the multiple sound collection devices, and the output unit outputs information corresponding to the frequency of the pronunciation based on the pronunciation position, which is the position where the pronunciation was made, identified using the relationship between the multiple arrival directions corresponding to the same sound arriving at the multiple sound collection devices.

3. A sound analysis device as described in claim 1 or 2, further comprising an identification unit that identifies an index corresponding to the diversity of the organisms by summing the frequencies of the pronunciations of each of the multiple types of organisms, and the output unit outputs the index as information corresponding to the frequencies of the pronunciations.

4. The sound analysis device according to claim 3, wherein the identification unit identifies the index by weighting the frequency of the sounds made by each of the plurality of types of organisms according to the rarity of the type.

5. The sound analysis device according to claim 3, wherein the identification unit identifies the indicator by weighting the frequency of the sounds made by each of the plurality of types of living organisms according to the danger of the type.

6. The sound analysis device described in claim 3, wherein the identification unit identifies the index for each of a plurality of sub-areas that make up the area by summing up the frequencies of the sounds made by each of a plurality of types of organisms in that sub-area, and the output unit outputs the index identified by the identification unit for each of the plurality of sub-areas in association with that sub-area.

7. The sound analysis device according to claim 1 or 2, wherein the output unit outputs information corresponding to the frequency of the pronunciation for each of a plurality of periods in association with the period.

8. The sound analysis device according to claim 1 or 2, wherein the acquisition unit acquires information indicating the state of each of the plurality of sound collection devices from the plurality of sound collection devices, and the output unit outputs information indicating a route to reach one or more of the plurality of sound collection devices whose state satisfies a predetermined condition.

9. The sound analysis device according to claim 8, wherein the output unit outputs information indicating the presence of a specified living thing in association with the route, on the condition that the acquisition unit acquires a sound emitted by the specified living thing within a specified range including the route.

10. A sound analysis device as described in claim 1 or 2, further comprising a generation unit that generates sound according to the information acquired by the acquisition unit, and the output unit outputs the sound generated by the generation unit from a sound output unit possessed by one or more of the plurality of sound collection devices.

11. The sound analysis device according to claim 10, wherein the generation unit generates a sound corresponding to the type of the organism that made the sound.

12. The sound analysis device described in claim 11, wherein the generation unit generates a sound that conveys a specific intention to the type of living thing that made the sound, and the output unit outputs the sound generated by the generation unit from the sound output unit of one or more sound collection devices among the multiple sound collection devices to which the sound emitted by the type of living thing has arrived.

13. A sound analysis method, executed by a processor, comprising: a step of acquiring information relating to sound acquired by a plurality of sound collection devices arranged in a predetermined area; and a step of outputting information corresponding to the frequency of sounds made by living organisms in the area based on the information acquired in the acquiring step.

14. A program that causes a processor to execute the steps of: acquiring information related to sounds captured by a plurality of sound collection devices arranged in a predetermined area; and outputting information corresponding to the frequency of sounds made by living organisms in the area based on the information acquired in the acquiring step.

Images