JP6138393B1 - Anomaly detection device, crime prevention system and anomaly detection method - Google Patents

Abstract

An anomaly detection apparatus that can easily set an anomaly detection for an object to be detected such as a window. An abnormality detection apparatus 100 includes a reference amplitude amount and a reference frequency, which are threshold information used for determining whether or not the state of the detected object is abnormal for each detected object of a window or a door. From the vibration sensor 200 based on the detected object information, the storage unit 20 for storing the detected object information, the sensor information acquiring unit 11 for acquiring the vibration information from the vibration sensor 200 for detecting the vibration information of the detected object. A processing unit 30 having a first detection mode for detecting that the detected object is abnormal when the amplitude amount of the acquired vibration information is equal to or greater than the reference amplitude amount and the frequency of the vibration information is equal to or greater than the reference frequency. . [Selection] Figure 1

Description

  The present invention relates to an abnormality detection device, a crime prevention system, and an abnormality detection method for detecting an abnormality of an object to be detected such as a window.

  Suspicious people are said to have the largest number of “window glass breaks” with more than 60% of intrusions into apartment buildings, detached houses, etc., and window glass is the most important point for crime prevention. ing.

  In order to prevent the intrusion of suspicious individuals, it is extremely effective to sign a crime prevention company with a crime prevention company and keep it in a 24-hour monitoring state. However, this measure has a cost problem, and buildings that cannot be contracted with crime prevention companies, such as ordinary houses and simple offices, are suspicious when they are vulnerable to crime prevention, such as when they are away or at midnight. It is extremely difficult to prevent an intrusion from a person's window.

  There are various techniques for “breaking window glass” such as “breaking” using a screwdriver, “breaking” using a bar, etc., and “burning” using a burner. In many cases, a method is used to create a fracture mark that is large enough to insert a hand by giving a certain impact to the window glass, and then insert the hand to unlock and open the window glass. It is. Therefore, it is possible to take crime prevention measures by detecting the situation when an impact is applied to the window glass.

  Patent Document 1 describes a glass breakage detection apparatus for destructive actions such as breakage that breaks glass by hitting it with a hammer or the like, and cracking by breaking with a flathead screwdriver.

JP 2011-187018 A

  The glass breakage detection apparatus of Patent Document 1 is not limited to ordinary float glass, but can also be laminated glass or the like, depending on the type, size, and thickness of the target glass. Therefore, it is necessary to perform circuit adjustment, circuit setting, etc. of the glass breakage detection apparatus, and there is a problem that a great deal of labor is required.

  The present invention is an invention for solving the above-described problems, and provides an abnormality detection device, a crime prevention system, and an abnormality detection method capable of accurately detecting abnormality with respect to objects to be detected having different materials and sizes. With the goal.

  In order to achieve the above object, the abnormality detection device of the present invention provides a reference amplitude which is threshold information used for determining whether or not the state of the detected object is abnormal for each detected object of a window or a door. A storage unit that stores the amount and the reference frequency as detected object information, a sensor information acquiring unit that acquires vibration information from a vibration sensor that detects vibration information of the detected object based on the detected object information, and a vibration sensor A processing unit having a first detection mode for detecting that the detected object is abnormal when the amplitude amount of the vibration information acquired from the reference amplitude amount is equal to or greater than the reference amplitude amount and the frequency of the vibration information is equal to or greater than the reference frequency. It is characterized by that. Other aspects of the present invention will be described in the embodiments described later.

  According to the present invention, it is possible to accurately detect abnormality of objects to be detected such as different materials and sizes.

It is a figure which shows the structure of the crime prevention system containing the abnormality detection apparatus which concerns on this embodiment. It is a figure which shows the structure which applied the crime prevention system to the apartment house. It is a figure which shows the installation position and monitoring target range of a sensor. It is a figure which shows a to-be-detected object database. It is a figure which shows an image information database. It is a figure which shows the method of the abnormality detection based on sensor information, (a) is a preliminary | backup detection mode, (b) (c) is a 1st detection mode time. It is a flowchart which shows the process of abnormality detection of preliminary | backup detection mode and 1st detection mode. It is a time chart which shows an example of abnormality detection of the 1st detection mode, when (a) shifts to the 1st detection mode, (b) is a case where abnormality is not detected in the 1st detection mode, (c) This is a case where an abnormality is detected in one detection mode. It is a flowchart which shows the process of abnormality detection of a 2nd detection mode. It is a figure which shows the structure which applied the crime prevention system to a detached house, (a) is an arrangement | positioning position of a sensor, (b) is an example in the case of monitoring one side of a house.

DESCRIPTION OF EMBODIMENTS Embodiments for carrying out the present invention will be described in detail with reference to the drawings as appropriate.
FIG. 1 is a diagram illustrating a configuration of a crime prevention system including an abnormality detection device according to the present embodiment. The security system SS (Security System) can easily set anomaly detection for an object to be detected such as a window and a door, and also includes an anomaly detection device 100 having a plurality of detection mode units, and a vibration sensor 200 for detecting vibration information. And an imaging device 300 that detects image information. The abnormality detection device 100 notifies the external terminal 400 that the detected object is abnormal while notifying the detected object with a buzzer or the like when the detected object is abnormal.

  The abnormality detection apparatus 100 includes an input unit 10, a storage unit 20, a processing unit 30, a communication unit 40 that communicates with an external terminal 400 via a network NW (for example, the Internet), a notification unit 50 such as a buzzer, and the like. The input unit 10 includes a sensor information acquisition unit 11 that acquires vibration information from the vibration sensor 200 and an image information acquisition unit 12 that acquires image information from the imaging apparatus 300.

  The storage unit 20 includes a detected object database 21, an image information database 22, a trigger amplitude amount 23, sensor information 24 acquired by the sensor information acquisition unit 11, image information 25 acquired by the image information acquisition unit 12, and related information 26. Is remembered. The related information 26 is, for example, information that associates identification information for each vibration sensor 200, position information of the vibration sensor 200, and a setting number (see FIG. 4) in the detected object database 21. A setting number may be registered in the vibration sensor 200.

  The detected object database 21 includes detected object information indicating the type, material, size, type, and the like of a frame for each detected object such as a window and a door, a reference amplitude amount corresponding to the detected object information, a reference frequency, The trigger amplitude amount is included. The reference amplitude amount, the reference frequency, and the like are threshold information used when determining whether or not the state of the detected object is abnormal.

  The image information database 22 includes reference image information acquired in advance for the monitoring target on which the vibration sensor 200 is installed, threshold information used when determining whether or not there is a suspicious person, and the like. The trigger amplitude amount 23 is used as a threshold value when the sensor information acquisition unit 11 changes the sensor information capturing period or excludes vibrations caused by wind or the like. Details of the detected object database 21 and the image information database 22 will be described later with reference to FIGS.

  The processing unit 30 includes a preliminary detection mode unit 31 for processing a preliminary detection mode based on sensor information, a first detection mode unit 32 for processing a first detection mode based on sensor information, and a second detection based on image information. It has the 2nd detection mode part 33 which processes a mode.

  The preliminary detection mode is a normal detection mode. When the vibration information is acquired, the vibration information is acquired from the vibration sensor 200 at the interval of the first reference time, and the amplitude amount of the vibration information is equal to or greater than the trigger amplitude amount 23. The first detection mode. The preliminary detection mode is a mode for providing an abnormality detection device with low power consumption and long battery life.

  The first detection mode is a mode in which the detected object is detected as abnormal when the amplitude amount of the vibration information acquired from the vibration sensor 200 is equal to or greater than the reference amplitude amount and the frequency of the vibration information is equal to or greater than the reference frequency. It is. The processing unit 30 shifts to the second detection mode when the detected object is abnormal in the first detection mode.

  In the second detection mode, the acquired image information is compared with the reference image information acquired in advance, and when the difference between the acquired image information and the reference image information exceeds a predetermined magnitude, This is the mode to detect. When the processing unit 30 shifts to the second detection mode, the processing unit 30 acquires the image information 25 from the imaging device 300 at intervals of a second reference time shorter than the first reference time, and digitizes the information by a predetermined method. Store in the storage unit 20.

  As the predetermined method, for example, a captured image is binarized (N-valued (N is a positive integer of 2 or more)) and is determined by a change in white area and black area. If the determination is based on the white area, if the difference in white between the initial image without the suspicious person M (see FIG. 3) and the image taken at the time of abnormality is equal to or greater than the white determination criterion or the black determination criterion, it is determined that there is a suspicious person M. . Alternatively, if the black area is determined, if the black difference between the initial image without the suspicious person M and the image taken at the time of abnormality is equal to or larger than the black determination criterion or the white determination criterion, it is determined that there is the suspicious person M.

  The processing unit 30 acquires the identification information given to the vibration sensor 200 that acquired the vibration information, and corresponds to the identification information if the imaging device 300 can perform pan / tilt when the mode is shifted to the second detection mode. A signal for setting the imaging direction of the imaging device 300 in the direction of the vibration sensor 200 to be output may be output to the imaging device 300. Of course, as long as the imaging apparatus 300 does not have a pan / tilt function, the imaging apparatus 300 may have a wide-angle lens and be fixed.

  When the processing unit 30 compares the reference image information digitized at the calibration time interval stored in the storage unit 20 and the image information obtained from the imaging apparatus 300 and digitized, the processing unit 30 deviates more than a predetermined size. In addition, the position correction signal may be output to the imaging apparatus 300 so that the magnitude of the deviation is within a predetermined value.

  The abnormality detection apparatus 100 is connected to the network NW via the communication unit 40 and can receive additional information or correction information of the detected object information from the external terminal 400.

  The vibration sensor 200 is installed for each detected object in the monitoring target range of the imaging device 300. Specifically, it is arranged at a predetermined position (near the crescent) on the indoor side of the window glass. The vibration sensor 200 includes a piezoelectric vibration sensor that can extract voltage signals of speed and acceleration, an electrodynamic vibration sensor that can extract voltage vibrations of speed, and a non-contact type eddy current displacement sensor. In the case of an eddy current displacement sensor, the object to be detected is a metal, so a metal sheet or the like may be attached to the window glass surface.

  The imaging device 300 is, for example, a CCD (Charge Coupled Device) camera, and is installed at a position that covers the monitoring target range. In addition, the imaging apparatus 300 images the monitoring target range with time, and outputs the captured image to the A / D converter as image information. The A / D converter is an electronic circuit that converts image information input as an analog signal into a digital signal and outputs the digital signal. When the imaging apparatus 300 is an analog signal output, an A / D converter may be included in the image information acquisition unit 12 or may be incorporated in the imaging apparatus 300.

  FIG. 2 is a diagram illustrating a configuration in which a security system is applied to an apartment house. FIG. 2 shows an example of a five-story apartment house. The second to fifth floors are residential areas and are composed of eight units. Each door is area 310. In the area 310, there are three sliding window glasses 390 (see FIG. 3) that serve as entrances and exits to the veranda, and one imaging device 300 monitors one house in the apartment house.

  FIG. 3 is a diagram illustrating the installation position of the vibration sensor and the monitoring target range. FIG. 3 shows an imaging range 320 on the veranda side in the imaging apparatus 300. It can be seen that the vibration sensors 200 are respectively disposed on the indoor window glass surface near the crescent of the sliding window glass 390. The imaging range 320 includes three sliding window glasses 390. The imaging range 320 is a range in which imaging can be performed once if the imaging device 300 does not have pan / tilt.

  FIG. 3 shows a case where there is a suspicious person M at the end of the veranda, and the details will be described later. However, in the second detection mode, the processing unit 30 differs between the initial screen and the imaging screen at the time of abnormality. Thus, the presence / absence of the suspicious person M is determined.

  FIG. 4 is a diagram showing the detected object database. The detected object database 21 includes identification information (setting No.), frame type, material, size, amplitude (may be speed or acceleration), capture cycle (or frequency) range, and the like. In the detected object database 21, 44 types of objects are registered. Frame types include aluminum, resin, wood, and steel. Examples of the material include float plate glass, mold plate glass, ground glass, mesh plate glass, wire plate glass, tempered glass, laminated glass, and multilayer glass. There are various sizes such as portrait, landscape, and thickness.

The amplitude includes a threshold value L1 used for detection determination in the preliminary detection mode, and a threshold value L2 (reference amplitude amount) used for detection determination in the first detection mode. As the capture period (capture time interval), there are a first reference time interval ΔT ₁ and a second reference time interval ΔT ₂ (<ΔT ₁ ).

  FIG. 5 is a diagram showing an image information database. The image information database 22 includes identification information (setting No.), an area of the imaging range 320, an abnormal area, a determination criterion, and the like. The area of the imaging range 320 includes a white area, a black area, and a total area. The judgment criteria include a white area and a black area threshold. The white area and the black area are preferably shown as a percentage of the total area. The abnormal area is a difference between an initial image captured in advance and a captured image.

FIGS. 6A and 6B are diagrams illustrating an abnormality detection method based on sensor information. FIG. 6A illustrates a preliminary detection mode, and FIGS. 6B and 6C illustrate a first detection mode. In the case of FIG. 6A, the processing unit 30 takes in the sensor information every interval ΔT ₁ of the first reference time, and moves to FIG. 6B when the threshold value L1 is reached. In FIG. 6A, ΣΔT ₁ indicates a combination of data for each ΔT ₁ . Note that ΔT ₁ is desirably equal to or less than a period of ½ frequency of the vibration waveform serving as an expected trigger.

For FIG. 6 (b), the processing unit 30 determines captures the sensor information at intervals [Delta] T ₂ of the second reference time, whether there is a case where equal to or greater than a threshold value L1. The reason for this processing is to exclude vibration caused by wind or the like. When it is equal to or greater than the threshold value L1, the process proceeds to FIG.

In FIG. 6 (c), when an impact corresponding to the destruction or destruction of the window glass is applied, a higher amplitude is generated at a higher frequency than in FIGS. 6 (a) and 6 (b). Thus, incorporation of the sensor information for each interval [Delta] T ₂ of the second reference time, if it becomes more than the threshold value L2, to detect the detection object is abnormal. As a detection method at a high frequency, for example, the sensor information 24 may be determined by comparing with the threshold value L2 via a high-pass filter.

FIG. 7 is a flowchart showing an abnormality detection process in the preliminary detection mode and the first detection mode. When the security monitoring is started, the processing unit 30 requests identification information from the vibration sensor 200, reads identification information (setting No.) of a return result from the vibration sensor 200 (step S71), and based on the identification information, An initial value is read based on the detected object database 21 (step S72). As initial values, there are a first reference time interval ΔT ₁ , a second reference time interval ΔT ₂ (<ΔT ₁ ), a threshold value L 1, and a threshold value L ₂ .

Processor 30 sets the read interval of the sensor information [Delta] T ₁ (step S73), after [Delta] T ₁ set in step S73, reads the data stored in the past [Delta] T ₁ hour every [Delta] T _1, the sensor information It is determined whether or not the sensor value L is greater than or equal to a threshold value L1 (L ≧ L1) (step S74). If the sensor value L is less than the threshold value L1 (step S74, No), the process returns to step S74, and if the sensor value L is greater than or equal to the threshold value L1 (step S74, Yes), the process proceeds to step S75.

In step S75, the processing unit 30 enters the first detection mode is set to read interval [Delta] T ₂ of the sensor information, after [Delta] T ₂ set in step S75, the data stored in the past [Delta] T ₂ hours every [Delta] T ₂ , And it is determined whether or not the sensor value L of the sensor information is not less than the threshold value L1 (L ≧ L1) within the predetermined time Ta (step S76). Here, the predetermined time Ta is a time for determining whether to return from the first detection mode to the preliminary detection mode. If the sensor value L is less than the threshold value L1 within the predetermined time Ta (step S76, No), the process returns to step S73, and if the sensor value L is greater than or equal to the threshold value L1 within the predetermined time Ta (step S76, Yes), step Proceed to S77.

  In step S77, the processing unit 30 performs a high-pass filter process, and determines whether or not the sensor value L of the sensor information is greater than or equal to a threshold value L2 (L ≧ L2) (step S78). If the sensor value L is less than the threshold value L2 (step S78, No), the process returns to step S76, and if the sensor value L is greater than or equal to the threshold value L2 (step S78, Yes), the process proceeds to step S79.

  In step S79, the processing unit 30 recognizes that the detected object is abnormal, and notifies the notification unit 50 of the abnormality. For example, an emergency sound may be generated from the notification unit 50. If there is an emergency sound, there is an intimidating effect on the suspicious person M. And the process part 30 transfers to 2nd detection mode (step S90). Details of the second detection mode will be described later with reference to FIG.

  FIG. 8 is a time chart showing an example of abnormality detection in the first detection mode. (A) shows a case where the first detection mode is entered, (b) shows a case where no abnormality is detected in the first detection mode, (c) ) Is a case where an abnormality is detected in the first detection mode.

For FIG. 8 (a), the processing unit 30, as a preliminary detection mode starts uptake at intervals [Delta] T ₁ of the first reference time is input period, the sensor value L of the sensor information threshold value L1 or more (L ≧ L1) a as (see time t1), the process proceeds to the first detection mode starts to capture at intervals [Delta] T ₂ of the second reference time. Within a predetermined time Ta, if the sensor value L is less than the threshold value L1, returns the input period in the interval [Delta] T ₁ of the first reference time, it continues to monitor.

  In the case of FIG.8 (b), although the process part 30 detected the sensor value L more than the threshold value L1 within predetermined time Ta, it has not detected the vibration which leads to destruction of a window glass. For example, this is an example in which vibrations such as wind continue. If vibration exceeding the threshold value L1 is detected within the first predetermined time Ta, the timer for measuring the predetermined time Ta is reset, and whether there is vibration that may break the window glass again within the predetermined time Ta. Continue monitoring.

  In the case of FIG.8 (c), the process part 30 detected the sensor value L more than threshold value L1 within predetermined time Ta, and the sensor value L of the sensor information was more than threshold value L2 in 1st detection mode. So it is certified abnormal. That is, there is a high possibility that the object to be detected is destroyed, and the mode shifts to the second detection mode in order to confirm the presence of the suspicious person M.

  FIG. 9 is a flowchart showing an abnormality detection process in the second detection mode. When the processing unit 30 receives that the detected object is abnormal in the first detection mode, the processing unit 30 requests the imaging apparatus 300 for identification information, and reads the identification information (setting No.) of the return result from the imaging apparatus 300. (Step S91), a criterion is read based on the image information database 22 (Step S92). As a determination criterion, there are a white area and a black area of a binarized white / black threshold.

  The processing unit 30 reads the initial area of the initial image that is the monitoring target range (step S93). The initial area is the white area, black area, and total area of the imaging range area of the binarized white / black image information database 22.

  The processing unit 30 acquires an imaging screen (step S94), and calculates the area (white area, black area) of the captured image (step S95). For example, when calculating the area, the captured image may be binarized to calculate the total value of white and black pixels, and the ratio to the total pixel value may be obtained.

  The processing unit 30 calculates a difference between the initial image and the captured image (step S96), and determines whether or not the difference is a determination criterion abnormality (step S97). If the difference is equal to or greater than the determination criterion (step S97, Yes), the processing unit 30 recognizes that there is a suspicious person M and notifies the external terminal 400 or the like (step S9A). On the other hand, if the difference is not greater than or equal to the determination criterion (No in step S97), the processing unit 30 determines whether or not the number of times greater than or equal to the determination criterion is greater than or equal to the predetermined number (step S98). If it is not the predetermined number or more (step S98, No), the process returns to step S94, and if it is the predetermined number or more (step S98, Yes), the processing unit 30 notifies the external terminal 400 or the like that the abnormality is not recognized ( Step S99), the process is terminated.

Specifically, FIG. 9 will be described.
The determination criterion is white area determination criterion WL, black plane line determination criterion BL, the initial image white area is W0, the black area is B0, and the white area of the captured image in the second detection mode is W1, black. The area is B1. The total area T0 = W0 + B0 = W1 + B1.
In step S96,
White difference ΔW = W0−W1, Black difference ΔB = B0−B1 (1)
It becomes.
In step S97, whether or not there is an abnormality is
| ΔW | ≧ WL or | ΔB | ≧ BL (2)
It can be determined by.
In addition, (1) Formula is
White difference ΔW = B1−B0, black difference ΔB = W1−W0 (3)
It is good.
Also, whether or not (2) is abnormal,
| ΔW | ≧ BL or | ΔB | ≧ WL (4)
It is good.
Here, if any one of the absolute values of the difference between the white area and the black area exceeds the criterion, it may be determined that there is a suspicious person M.

  According to the processing in the second detection mode in FIG. 9, it is further certain whether or not the determination of abnormality in the first detection mode is reliable.

  FIG. 10 is a diagram illustrating a configuration in which a crime prevention system is applied to a detached house, where (a) is an example of a sensor arrangement position, and (b) is an example in the case of monitoring a detached house. . FIG. 10A shows an example of a two-story house, in which a vibration sensor 200 is provided for windows of different sizes and materials. FIG. 10B shows an example in which the security system SS is applied to a one-story detached house. One side of the detached house has three windows WD1, WD2, and WD3. An aluminum frame ground glass is applied to the window WD1, and an aluminum frame float glass is applied to the window WD2. On the other hand, the window WD3 has the same frame type and window material as the window WD2, but has a different size.

The abnormality detection device 100 includes a determination unit (channel) for the preliminary detection mode and the first detection mode for each of the plurality of vibration sensors 200, and from the sensor information acquisition unit 11 every interval ΔT ₂ in the first detection mode. The sensor information is determined by a threshold value L2 for each different window through a high-pass filter.

  The high-pass filter can be handled by the same thing except for a special window. The reason is that when the glass is large, a thick plate is used, and when the glass is small, a thin plate is used. Therefore, the destruction frequency is substantially the same, and the destruction frequency is further increased when the plate thickness is thick. Therefore, the high-pass filter can be handled by the same. However, the larger the amount of amplitude, the larger the amount of glass and the smaller the amount of glass. Therefore, the final destruction is determined by the amplitude amount. However, in this case, imaging in the second detection mode has been started, and it is possible to verify in the second detection mode whether or not the detection that there is an abnormality in the first detection mode is correct. it can. Note that the special window for the above has a wooden or hybrid frame, and this can be dealt with by preparing a high-pass filter set at a different frequency.

  In FIG. 10A, the vibration sensor 200 is disposed on the window. However, the crime prevention system SS can be applied to any place where a suspicious person M may enter, such as an entrance door or a doorway door. .

  According to the present embodiment, it is possible to accurately grasp whether or not the detected object is abnormal in the first detection mode even if the detected object is of a different material or size. When an abnormality is determined in the first detection mode, the presence or absence of a suspicious person can be determined in the second detection mode.

  It should be noted that the present invention is not limited to the above-described embodiments, and various other application examples and modifications can be taken without departing from the gist of the present invention. For example, in FIG. 2, the imaging range 320 is imaged by one imaging device 300, and the presence or absence of the suspicious person M is determined in the second detection mode. When the imaging apparatus 300 is a network camera capable of remote pan / tilt, an initial image may be captured for each window, and the presence or absence of the suspicious person M may be determined according to the determination criterion for each window. In this case, when it is detected that there is an abnormality in the first detection mode, the network camera may be turned in the pan direction and the tilt direction based on the window identification information and the position information indicating that there is an abnormality. Further, since it is a network camera, in FIG. 1, the imaging apparatus 300 can be connected to the network NW and can acquire image information via the network NW. Further, when there is a suspicious person M in the second detection mode, the abnormality detection apparatus 100 can easily determine whether or not the user is a suspicious person by transmitting an image of the suspicious person M to the external terminal 400. Can do.

DESCRIPTION OF SYMBOLS 10 Input part 11 Sensor information acquisition part 12 Image information acquisition part 20 Memory | storage part 21 Detected object database 22 Image information database 23 Trigger amplitude amount 24 Sensor information 25 Image information 26 Related information 30 Processing part 31 Preliminary detection mode part 32 1st detection Mode unit 33 Second detection mode unit 40 Communication unit 50 Notification unit 100 Abnormality detection device 200 Vibration sensor 300 Imaging device 400 External terminal NW network SS Security system

Claims (10)

A memory that stores, as detected object information, a reference amplitude amount and a reference frequency that are threshold information used when determining whether or not the detected object state is abnormal for each detected object of a window or a door. And
A sensor information acquisition unit that acquires the vibration information from a vibration sensor that detects vibration information of the object to be detected;
Based on the detected object information, when the amplitude amount of the vibration information acquired from the vibration sensor is not less than the reference amplitude amount and the frequency of the vibration information is not less than the reference frequency, the detected object is An abnormality detection device comprising: a processing unit having a first detection mode for detecting an abnormality. The abnormality detection device further includes an image information acquisition unit that acquires image information from an imaging device that images the object to be detected.
The processing unit compares the acquired image information with reference image information acquired in advance, and when the difference between the acquired image information and the reference image information is a predetermined size or more, A second detection mode for detecting an abnormality,
The abnormality detection device according to claim 1, wherein the processing unit shifts to the second detection mode when the detected object is abnormal in the first detection mode. The storage unit stores a trigger amplitude amount that is threshold information used when determining whether or not the vibration is caused by wind for each detected object,
The processing unit obtains the vibration information from the vibration sensor at an interval of a first reference time when obtaining the vibration information, and the amplitude amount of the vibration information is equal to or greater than the trigger amplitude amount. The mode is shifted to the first detection mode, the vibration information is acquired from the vibration sensor at intervals of a second reference time smaller than the first reference time, and the acquired vibration information is stored in the storage unit. The abnormality detection device according to claim 1 or 2. When the processing unit shifts to the second detection mode, the processing unit acquires the image information from the imaging device at an interval of the second reference time, and quantifies the information by a predetermined method and stores it in the storage unit. The abnormality detection device according to claim 3, which is cited from claim 2. The storage unit stores identification information and position information of the vibration sensor in association with each other.
The processing unit acquires the identification information given to the vibration sensor that has acquired the vibration information, and when the processing unit shifts to the second detection mode, the imaging unit captures an image in the vibration sensor direction corresponding to the identification information. The abnormality detection device according to claim 2, wherein a signal for setting a direction is output to the imaging device. When the processing unit compares the reference image information digitized at the calibration time interval stored in the storage unit and the digitized image information acquired from the imaging device and deviates more than a predetermined size The abnormality detection device according to claim 2, wherein a position correction signal is output to the imaging device so that the magnitude of the deviation is within a predetermined value. In the storage unit, the detected object information indicating the type, material, shape, and size of the detected object that is the detected object, the reference amplitude amount corresponding to the detected object information, the reference frequency, The abnormality detection device according to claim 3, wherein a trigger amplitude amount is stored. An abnormality detection device according to any one of claims 2 to 7,
A vibration sensor for detecting vibration information;
An imaging device for detecting image information;
When the detected object is abnormal, the abnormality detection device transmits to the external terminal that the detected object is abnormal. The crime prevention system according to claim 8, wherein the abnormality detection device is connected to the Internet and receives additional information or correction information of the detected object information from the external terminal. A sensor information acquisition unit that acquires the vibration information from a vibration sensor that detects vibration information of a detection object of a window or a door; an image information acquisition unit that acquires image information from an imaging device that images the detection object; An abnormality detection method for detecting an abnormality of the detected object using an abnormality detection device including a processing unit and a storage unit,
In the storage unit, for each detected object, a reference amplitude amount and a reference frequency, which are threshold information used when determining whether or not the state of the detected object is abnormal, are detected object information. Remembered,
The processor is
Based on the detected object information, if the amplitude amount of the vibration information acquired from the vibration sensor is greater than or equal to a reference amplitude amount and the frequency of the vibration information is equal to or greater than the reference frequency, the detected object is abnormal. A first detection process to detect;
When the detected object is abnormal in the first detection process, the acquired image information is compared with reference image information acquired in advance, and the difference between the acquired image information and the reference image information And performing a second detection process in which the detected object is detected as abnormal when the value becomes equal to or greater than a predetermined size.

Images