Patent Application
20230401941
The present disclosure relates to a monitoring system, a monitoring apparatus, a monitoring method, and a computer readable medium.
In related techniques, an abnormality in a monitoring target such as a fence has been monitored by a monitoring person in a monitoring room monitoring camera images of a plurality of cameras. For example, when the monitoring person determines that there is a suspicious point in the monitoring target, he/she turns the orientation of the camera towards the monitoring target and controls the camera to zoom in so as to detect an abnormality in the monitoring target. However, it may take a lot of time for a human to detect an abnormality in a monitoring target, which delay may cause a large increase in the cost of finding and handling the abnormality.
Thus, recently, a system for monitoring an abnormality in a monitoring target using an optical fiber has been proposed (e.g., Patent Literatures 1 to 3).
In the technique described in Patent Literatures 1 and 2, an FBG (Fiber Bragg Grating) optical fiber is laid on a fence and an OTDR (Optical Time-Domain Reflectometry) optical fiber is also laid on the fence. An intruder attempting to climb over the fence is detected by an FBG intrusion detector, and an intruder attempting to break the fence is detected by an OTDR intrusion detector. An ITV (Industrial Television) monitoring system provided in a monitoring room turns a shooting direction of an ITV camera to a detection location based on detection location information included in an intrusion detection signal from the intrusion detector.
In the technique described in Patent Literature 3, a plurality of FBG optical fibers are laid on a fence, and a pulse signal is output when an amount of a reflected wave shift of reflected waves generated in the optical fibers exceeds a predetermined threshold. Further, a table is prepared, in which events occurring in the fence classified according to the number of vibrations and fluctuation amount of the fence are associated with delays in times when pulse signals are generated and generation frequencies of the pulse signals in the respective plurality of optical fibers. Then, delays in times when pulse signals are generated and generation frequency of the pulse signals in the respective plurality of optical fibers are checked against the above-mentioned table in order to detect an event occurring in the fence.
However, the techniques described in Patent Literatures 1 to 3 have a problem that it is necessary to use a special optical fiber called an FBG optical fiber. Further, the techniques described in Patent Literatures 1 and 2 have a problem that it is necessary to use an OTDR optical fiber together with an FBG optical fiber. Furthermore, the technique described in Patent Literature 3 has a problem that it is necessary to lay a plurality of optical fibers and receive reflected waves from the plurality of optical fibers.
Thus, an object of the present disclosure is to provide a monitoring system, a monitoring apparatus, a monitoring method, and a computer readable medium which solve the above-described problems and can detect a state of a monitoring target without using a special structure for detecting the state of the monitoring target.
In an example aspect, a monitoring system includes:
In another example aspect, a monitoring apparatus includes:
In another example aspect, a monitoring method performed by a monitoring apparatus includes:
In another example aspect, a non-transitory computer readable medium stores a program causing a computer to execute:
According to the above example aspects, it is possible to solve the above-described problems and achieve an effect that a state of a monitoring target can be detected without using a special structure for detecting the state of the monitoring target.
Hereinafter, an embodiment of the present disclosure will be described with reference to the drawings. In the embodiment described below, as an example, a monitoring target to be monitored is described as a fence, but the monitoring target is not limited to a fence.
First, a configuration of a monitoring system according to this embodiment will be described with reference to
As shown in
The optical fiber cable 20 is formed by covering one or more optical fibers. The optical fiber cable 20 is laid on the fences 10 and buried in the ground along the fences 10. Specifically, the optical fiber cable 20 extends from the optical fiber detection unit 31 along the fences 10 and is turned back at a turning point and returns to the optical fiber detection unit 31. A part of the optical fiber cable 20 between the optical fiber detection unit 31 and the turning point is laid on the fences 10, and the other part of the optical fiber cable 20 is buried in the ground along the fences 10. However, the method of laying and burying the optical fiber cable 20 shown in
The camera 40 photographs an area where the fences 10 are installed. The camera 40 is implemented by, for example, a fixed camera, a PTZ (Pan Tilt Zoom) camera, or the like.
The monitoring system according to this embodiment monitors the fences 10 and its surroundings using an optical fiber sensing technique that uses an optical fiber as a sensor.
Specifically, the optical fiber detection unit 31 makes pulsed light incident on at least one optical fiber included in the optical fiber cable 20. Then, as the pulsed light is transmitted through the optical fiber in the direction of the fences 10, backscattered light is generated at each transmission distance. This backscattered light returns to the optical fiber detection unit 31 via the above-mentioned same optical fiber as the one through which the pulsed light is transmitted.
At this time, the optical fiber detection unit 31 makes the pulsed light incident in the clockwise direction and receives the backscattered light from this pulsed light in the clockwise direction and also makes the pulsed light incident in the counterclockwise direction and receives the backscattered light from this pulsed light in the counterclockwise direction. Thus, the optical fiber detection unit 31 receives the backscattered light from two directions.
Here, the fence 10 vibrates when an event such as a person grabbing and shaking the fence 10 occurs, and the vibration of the fence 10 is transmitted to the optical fiber. The vibration pattern of the vibration of the fence 10 transmitted to the optical fiber is a dynamically fluctuating pattern and differs according to the type of an event occurring in the fence 10 and its surroundings. In the first embodiment, for example, the following events are assumed as predetermined events that occur in the fences 10 and its surroundings.
Thus, the backscattered light received from the optical fiber by the optical fiber detection unit 31 includes a pattern corresponding to the state of the fence 10 and its surroundings, i.e., a pattern corresponding to an event occurring in the fence 10 and its surroundings. Therefore, in this embodiment, the state of the fence 10 and its surroundings is detected by the method described below by using the fact that the pattern corresponding to the state of the fence 10 and its surroundings is included in the backscattered light. Specifically, a predetermined event occurring in the fence 10 and its surroundings is detected.
The optical fiber detection unit 31 can identify the location of the fence 10 in which this backscattered light is generated based on a time difference between a time when the pulsed light is incident on the optical fiber and a time when the backscattered light is received from this optical fiber. Further, in this embodiment, as described above, the fences 10 are composed of a plurality of fences 10 connected to each other. Therefore, as shown in
Thus, the optical fiber detection unit 31 can generate, for example, as sensing data vibration data as shown in
In the example shown in
Thus, in this embodiment, the control unit 32 performs machine learning (e.g., deep learning) on the vibration pattern when a predetermined event is occurring in the fence 10 and its surroundings and detects whether a predetermined event is occurring in the fence 10 and its surroundings using a result of the machine learning (initial training model).
First, a method of the machine learning will be described with reference to
As shown in
Next, the control unit 32 checks the vibration patterns against the supervised data and classifies the vibration patterns (Step S3) and performs supervised learning (Step S4). By doing so, an initial training model is obtained (Step S5). When a vibration pattern corresponding to an event occurring in the fence 10 and its surroundings is input, this initial training model outputs a predetermined event that may be applicable if there is a possibility that this event may correspond to any of the predetermined event. Alternatively, this initial training model may output, together with a predetermined event that may be applicable, confidence at which this predetermined event occurs.
Next, a method of determining whether a predetermined event occurring is detected in the fence 10 and its surroundings will be described.
In this case, the control unit 32 first acquires a vibration pattern corresponding to an event occurring in the fence 10 and its surroundings from the optical fiber detection unit 31. Next, the control unit 32 inputs this vibration pattern to the initial training model. By doing so, since the control unit 32 can obtain a predetermined event that may be applicable as a result of the output from the initial training model, it detects that a predetermined event is occurring. Moreover, when the control unit 32 obtains confidence together with a predetermined event that may be applicable as a result of the output from the initial training model, it may determine that the predetermined event occurring is detected if the confidence is more than or equal to a threshold.
As described above, in this embodiment, the vibration pattern when a predetermined event is occurring in the fence 10 and its surroundings is machine-learned, and a result of the machine learning is used to detect a predetermined event occurring in the fence 10 and its surroundings.
It may be difficult in an analysis by a human to extract, from data, features for detecting an event occurring in the fence 10 and its surroundings. In this embodiment, by building a training model from a large number of patterns, it is possible to detect a predetermined event occurring in the fence 10 and its surroundings with high accuracy even when it is difficult in an analysis by a human to do so.
Note that in the machine learning according to this embodiment, in the initial state, a training model may be generated based on two or more pieces of supervised data. In addition, this training model may be made to newly learn a newly detected pattern. At this time, a specific condition for detecting a predetermined event occurring in the fence 10 and its surroundings may be adjusted based on the new training model.
Further, as shown in
As shown in
Additionally, the control unit 32 may control two or more cameras 40 which photograph an area including the fence 10 in which the predetermined event is detected among the plurality of cameras 40. In this case, the function may be divided for each camera 40. For example, at least one of the two or more cameras 40 may photograph a face of a person present in the above-mentioned area, so that the photographed face image is used for face authentication, while another at least one of the two or more cameras 40 may photograph the above-mentioned entire area, so that the photographed image is used for monitoring a behavior of a person or an animal present in the above-mentioned area. Moreover, the two or more cameras 40 may photograph the area with different angles. Furthermore, at least one of the two or more cameras 40 may perform photographing to complement photographing of another camera 40. For example, when there is a blind spot that cannot be photographed by the other camera 40 in the above-mentioned area, the at least one camera 40 may photograph the blind spot. When there is a dark area at night, the control unit 32 may further control another camera 40 with a night vision function such as an infrared camera instead of the camera 40 for photographing the area or control the camera 40 to switch to a night vision mode. As shown in
The control unit 32 may further control the display unit 50 to display, for example, a sensing data image indicating sensing data generated by the optical fiber detection unit 31 and a camera image of the camera 40 which photographs an area including the fence 10 in which a predetermined event is detected based on the sensing data. However, a content of the display is not limited.
Next, a hardware configuration of a computer 60 that implements the monitoring apparatus 30 will be described with reference to
As shown in
The processor 601 is an arithmetic processing apparatus such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit). The memory 602 is a memory such as a RAM (Random Access Memory) or a ROM (Read Only Memory). The storage 603 is a storage apparatus such as an HDD (Hard Disk Drive), an SSD (Solid State Drive), or a memory card. The storage 603 may be a memory such as a RAM or a ROM.
The storage 603 stores programs that implement the functions of the optical fiber detection unit 31 and the control unit 32 included in the monitoring apparatus 30. The processor 601 implements the functions of the optical fiber detection unit 31 and the control unit 32 by executing the programs. Here, the processor 601 may execute these programs after reading them into the memory 602 or without reading them into the memory 602. The memory 602 and the storage 603 also play a role to store information and data held by the optical fiber detection unit 31 and the control unit 32.
The above programs can be stored and provided to a computer (including the computer 60) using any type of non-transitory computer readable media. Non-transitory computer readable media include any type of tangible storage media. Examples of non-transitory computer readable media include magnetic storage media (such as floppy disks, magnetic tapes, hard disk drives, etc.), optical magnetic storage media (e.g. magneto-optical disks), CD-ROM (Compact Disc-Read Only Memory), CD-R (CD-Recordable), CD-R/W (CD-ReWritable), and semiconductor memories (such as mask ROM, PROM (Programmable ROM), EPROM (Erasable PROM), flash ROM, RAM (Random Access Memory), etc.). The program may be provided to a computer using any type of transitory computer readable media. Examples of transitory computer readable media include electric signals, optical signals, and electromagnetic waves. Transitory computer readable media can provide the program to a computer via a wired communication line (e.g. electric wires, and optical fibers) or a wireless communication line.
The input/output interface 604 is connected to a display apparatus 6041, an input apparatus 6042, and so on. The display apparatus 6041 displays a screen corresponding to drawing data processed by the processor 601, such as an LCD (Liquid Crystal Display) or a CRT (Cathode Ray Tube) display. The input apparatus 6042 receives an operator's operation input and is, for example, a keyboard, a mouse, and a touch sensor. The display apparatus 6041 and the input apparatus 6042 may be combined and implemented as a touch panel. Note that the computer 60 may include a sensor (not shown) including a distributed vibration sensor, and the sensor may be connected to the input/output interface 604. The display unit 50 may be connected to the input/output interface 604.
The communication interface 605 transmits data to and receives data from an external apparatus. For example, the communication interface 605 communicates with an external apparatus via a wired communication path or a wireless communication path.
Hereinafter, an operation of the monitoring system according to this embodiment will be described. Here, an operation flow of the monitoring system according to this embodiment will be described with reference to
As shown in
Next, the optical fiber detection unit 31 receives backscattered light including a pattern corresponding to the state of the fence 10 and its surroundings from the same optical fiber as the optical fiber on which the pulsed light is incident and detects a pattern corresponding to the state of the fence 10 and its surroundings from the received backscattered light (Step S12). Specifically, the optical fiber detection unit 31 detects a pattern corresponding to an event occurring in the fence 10 and its surroundings. At this time, as described above, the optical fiber detection unit 31 detects strength of a vibration, a vibration location, a transition of fluctuation in the number of vibrations, and the like caused by the event occurring in the fence 10 and its surroundings, thereby detecting a dynamic fluctuation pattern of the fence 10 and its surroundings. Alternatively, the optical fiber detection unit 31 may further detect a dynamic fluctuation pattern in a sound and a temperature, thereby detecting a complex unique pattern in the fence 10 and its surroundings.
After that, the control unit 32 detects the state of the fence 10 and its surroundings based on the pattern detected by the optical fiber detection unit 31 (Step S13). Specifically, it is detected whether a predetermined event is occurring in the fence 10 and its surroundings. At this time, the control unit 32 may detect whether a predetermined event is occurring in the fence 10 and its surroundings by the above-described method of the machine learning.
As described above, according to this embodiment, backscattered light (light signal) including a pattern corresponding to the state of the fence 10 and its surroundings is received from at least one optical fiber included in the optical fiber cable 20, the pattern is detected from the received backscattered light, and the state of the fence 10 and its surroundings is detected based on the detected pattern. In this manner, since the state of the fence 10 and its surroundings is detected based on the pattern included in the backscattered light, it is not necessary to use a special optical fiber called an FBG optical fiber like in Patent Literatures 1 to 3, to use an OTDR optical fiber together with an FBG optical fiber like in Patent Literatures 1 and 2, and to lay a plurality of optical fibers like in Patent Literature 3. Therefore, the state of the fence 10 and its surroundings can be detected without using a special structure for detecting the state of the fence 10 and its surroundings. Moreover, the monitoring system can be built for a low cost, because a special structure is not required.
Further, according to this embodiment, the state of the fence 10 and its surroundings is detected based on the pattern corresponding to the state of the fence 10 and its surroundings included in the backscattered light. That is, instead of separating the states of the fence 10 according to the strength of vibrations and the number of vibrations like in Patent Literature 3 (e.g., the states are identified according to whether or not the vibration is large and whether or not the number of vibrations is high), in this embodiment, the state of the fence 10 and its surroundings is detected by a dynamic pattern analysis on the strength of the vibration and the number of vibrations (e.g., transition of a change in the strength of vibrations). This enables an accurate detection of the state of the fence 10 and its surroundings.
Further, according to this embodiment, the state of the fence 10 and its surroundings is detected based on a pattern corresponding to the state of the fence 10 and its surroundings included in backscattered light. This enables a clear isolation of a slight change such as an event in the surroundings of the fence 10 not involving contact with the fence 10 from other noise components including wind by detecting a dynamic pattern fluctuation, thereby accurately detecting a state in which this event is occurring.
Further, the FBG optical fiber used in Patent Literatures 1 to 3 has a configuration in which grating parts are provided at constant intervals, and in which an intruder is detected at the “points” where the grating parts are provided. On the contrary, in this embodiment, since a dynamic fluctuation pattern such as the vibration pattern in
Moreover, according to this embodiment, the optical fiber sensing technique which uses an optical fiber as a sensor is used. Thus, this embodiment has advantages such that an optical fiber is not affected by electromagnetic noises, it is not necessary to supply power to a sensor, and it is possible to achieve excellent environmental resistance and easy maintenance.
In the above-described embodiment, an example in which the monitoring target is the fences 10 (and its surroundings) has been described, but the monitoring target is not limited to the fences 10 (and its surroundings). First, the installation site of the monitoring target may be an airport, a port, a plant, a nursing facility, a company building, a border, a nursery, a home, or the like. The monitoring target may be, a wall, a pipeline, a utility pole, a civil engineering structure, a floor, etc. in addition to a fence. Further, a laying or burying site of the optical fiber cable 20 when the monitoring target is monitored may be a wall, a pipeline, a utility pole, a civil engineering structure, a floor, etc., in addition to a fence and underground. For example, when the fence 10 installed in a nursing facility is to be monitored, examples of a predetermined event that could occur in the fence 10 include a person hitting the fence 10, a person leaning against the fence 10 due to injury or the like, and a person climbing over the fence 10 to escape.
In the above-described embodiment, it has been described that the fence 10 vibrates when a predetermined event occurs. However, when such an event occurs, a sound, a temperature, a strain, a stress, and the like change in the fence 10, and these changes are transmitted to the optical fiber. The patterns of a sound, a temperature, a strain, a stress, and the like are also dynamically fluctuating patterns and differ according to the type of an event occurring in the fence 10. For this reason, the optical fiber detection unit 31 may use a Distributed Acoustic Sensor, a Distributed Temperature Sensor, etc. in addition to a distributed vibration sensor to detect a change in a vibration, a sound, a temperature, strain, and stress, etc. and generate sensing data. Then, the control unit 32 may detect an event occurring in the fence 10 based on the sensing data in which the change in the vibration, the sound, the temperature, the strain, and stress, etc. is reflected. This further improves a detection accuracy.
In the above-described embodiment, when a predetermined event is occurring in the fence 10, the control unit 32 controls the angle, zoom magnification, and so on of the camera 40 which photographs an area including this fence 10. However, the control unit 32 may continue to control the camera 40 even after a predetermined event has occurred. For example, the control unit 32 may control the camera 40 to track a person, an animal, a car, and the like present in the above-mentioned area. Moreover, when a person wandering around the fence 10 leaves an object such as a suspicious object, the control unit 32 may control one camera 40 to photograph the object and another camera 40 to track the person.
Moreover, the optical fiber detection unit 31 and the control unit 32 of the monitoring apparatus 30 may be provided separately from each other. For example, the optical fiber detection unit 31 may be provided in a communication carrier station, and the monitoring apparatus 30 including the control unit 32 may be provided outside the communication carrier station.
In the above-described embodiment, only one optical fiber detection unit 31 is provided and the optical fiber cable 20 is occupied. However, the present disclosure is not limited to this.
For example, the optical fiber detection unit 31 may be provided in a communication carrier station, and the optical fiber cable 20 may be shared between existing communication equipment provided inside the communication carrier station and the optical fiber detection unit 31.
Alternatively, one optical fiber detection unit 31 may be provided in each of the plurality of communication carrier stations, and the detection unit fiber cable 20 may be shared between the plurality of optical fiber detection units 31 provided in the plurality of respective communication carrier stations.
Further alternatively, a plurality of optical fiber detection units 31 may be provided in one communication carrier station, and the optical fiber cable 20 may be shared between the plurality of optical fiber detection units 31.
Although the present disclosure has been described with reference to the embodiments, the present disclosure is not limited to the above-described embodiments. Various changes that can be understood by those skilled in the art can be made to the configurations and details of the present disclosure within the scope of the present disclosure.
The whole or part of the above embodiments can be described as, but not limited to, the following supplementary notes.
A monitoring system comprising:
The monitoring system according to Supplementary note 1, wherein
The monitoring system according to Supplementary note 1 or 2, further comprising a plurality of cameras, wherein
The monitoring system according to Supplementary note 3, wherein
The monitoring system according to Supplementary note 4, wherein
The monitoring system according to any one of Supplementary notes 1 to 5, wherein
A monitoring apparatus comprising:
The monitoring apparatus according to Supplementary note 7, wherein
The monitoring apparatus according to Supplementary note 7 or 8, wherein
The monitoring apparatus according to Supplementary note 9, wherein
The monitoring apparatus according to Supplementary note 10, wherein
The monitoring apparatus according to any one of Supplementary notes 7 to 11, wherein
A monitoring method performed by a monitoring apparatus, the monitoring method comprising:
A non-transitory computer readable storing a program causing a computer to execute:
This application is a continuation of U.S. application Ser. No. 17/289,462, filed Apr. 28, 2021 which is a National Stage of International Application No. PCT/JP2018/041348 filed Nov. 7, 2018, the entire contents of all of which are incorporated herein by reference.
| Number | Date | Country | |
|---|---|---|---|
| Parent | 17289462 | Apr 2021 | US |
| Child | 18234173 | US |
