JP2012004720A - Monitor, and monitoring system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a monitor of a monitoring camera using an IP camera or a network camera which is expected to be widely used in the future, in which a network load can be reduced in response with the increase of data size, and by which can reduce occasions of failing to detect abnormal behavior of figures taken by the camera.SOLUTION: A threshold to detect abnormality is set and only video data that has detected abnormal behavior is transmitted from a monitoring camera to an abnormality monitoring device. It is desirable to detect behavior close to the threshold, such as behavior as a sign of abnormal behavior. To satisfy this requirement, an abnormality comparison part determines abnormal behavior after detecting a specific figure from a video taken, and a threshold generation part changes a threshold for abnormal determination when a value of the determination for abnormal behavior by the figure falls within the specific range from the threshold to determine abnormality, for a predetermined number of times.

Description

  The present invention relates to a monitoring device and a monitoring system, and more particularly, to a monitoring device and a monitoring system that reduce detection failure of abnormal behavior.

  There exists patent document 1 as a background of this technical field. In this publication, “a method for setting an image monitoring area in which an abnormality monitoring condition such as a monitoring area setting of an abnormality monitoring apparatus using a video signal of a monitoring camera and an arithmetic function setting can be easily set from the installation site of the monitoring camera, and In response to the problem “Proposal of an abnormality monitoring system”, “a monitoring camera 1 for imaging a monitoring target, image data generation means 10 for converting and generating a video signal into digital image data, a monitoring area setting unit 11 for setting a monitoring area,” An image data calculation unit 12 that calculates image data in the monitoring area, an operation function setting unit 13 that sets an operation function used by the image data calculation unit, and an abnormality determination unit that compares the image calculation result with a predetermined threshold value to determine an abnormality. 14. An abnormality monitoring device 2 having a threshold selection unit 15 for selecting a threshold and an abnormality notification unit 16 for notifying and outputting an abnormality when the abnormality determination unit 14 determines that an abnormality has occurred. , Which describes the use of solutions that region instruction pointer "for instructing the monitored area.

Japanese Patent Laid-Open No. 2001-6058

  In IP cameras and network cameras that are expected to become popular in the future, it is necessary to reduce the network load by increasing the data size due to high-definition video data. It is necessary to transmit only data from the monitoring camera to the abnormality monitoring device. For that purpose, by having the abnormality monitoring function described in Patent Document 1 in the surveillance camera body, it is possible to transmit only the video determined to be abnormal and improve the network load.

However, although the abnormality monitoring device described in Patent Document 1 includes the threshold selection unit 15, the threshold set by the selection unit 15 is calculated and output as an abnormality determination threshold using image data obtained by imaging a simulated intruder. There is. Although it is possible to detect abnormal behavior that can be determined to be abnormal from the video with this method, it detects behavior that is a precursor to abnormal behavior, such as an intruder that is below the threshold for abnormal behavior or behavior immediately before the abnormal behavior occurs. I can't.
The objective of this invention is providing the monitoring apparatus and monitoring system which reduced the detection omission of abnormal action in view of an above described subject.

To achieve the above object, the present invention is a monitoring device for monitoring an abnormal behavior of a subject,
A personal authentication unit for identifying a person included in the video supplied with an image of the subject; an abnormal behavior recognizing unit for detecting an abnormal degree of the behavior of the person included in the video supplied with the image of the subject; The threshold generation unit that determines the threshold value of the behavioral abnormality level according to the person specified by the personal authentication unit, and the behavioral abnormality level detected by the abnormal behavior recognition unit and the threshold value set by the threshold value generation unit are compared. An abnormality level comparing unit for determining presence / absence of abnormal behavior of the person, and an abnormality notification unit for notifying an external device of an abnormality when the abnormality level comparing unit determines that the person has abnormal behavior. Have
When the abnormality degree of the behavior detected by the abnormal behavior recognition unit is smaller than the threshold value generated by the threshold value generation unit within a predetermined range, and / or continues for a predetermined period In this case, the threshold value generation unit reduces the threshold value by a predetermined amount.

The present invention also provides a monitoring system for monitoring abnormal behavior of a subject,
An imaging unit that images the subject and converts it into a digital signal, an image processing unit that converts the bit rate of the digital signal supplied from the imaging unit, and a digital signal after the bit rate conversion supplied from the image processing unit An external interface unit that sends out the outside of the monitoring system, and an abnormality detection unit that controls a bit rate conversion operation in the image processing unit,
The abnormality detection unit
A personal authentication unit that identifies a person included in the video image supplied by the image captured by the image capturing unit, and a degree of abnormality in the behavior of the person included in the video image provided by the image captured by the image capturing unit. An abnormal behavior recognizing unit to detect; a threshold generating unit for determining a threshold of the abnormal degree of the behavior according to the person specified by the personal authentication unit; and the abnormal degree of the behavior detected by the abnormal behavior recognizing unit and the threshold generation An abnormality level comparison unit that compares the threshold values determined by the unit and determines the presence or absence of abnormal behavior of the person,
When the abnormality degree of the behavior detected by the abnormal behavior recognition unit is smaller than the threshold value generated by the threshold value generation unit within a predetermined range, and / or continues for a predetermined period In this case, the threshold generation unit reduces the threshold by a predetermined amount,
And controlling the image processing unit to increase the bit rate of the digital signal after bit rate conversion in the image processing unit when the abnormality level comparing unit determines that the person has an abnormal behavior. It is characterized by.

  ADVANTAGE OF THE INVENTION According to this invention, there exists an effect that the monitoring apparatus and monitoring system which reduced the detection omission of abnormal action can be provided, and it can contribute to the improvement of the basic performance.

It is a block diagram of the monitoring system which is one Example of this invention. It is a figure explaining the threshold value which determines the abnormality degree of Example 1. FIG. It is a figure explaining the threshold value which determines the abnormality degree of Example 1. FIG. It is a figure explaining the threshold value which determines the abnormality degree of Example 1. FIG. It is a figure explaining the threshold value which determines the abnormal degree of Example 2. FIG. It is a figure explaining the threshold value which determines the abnormal degree of Example 2. FIG. It is a block diagram of the monitoring system of Example 3. It is a block diagram of the monitoring system of Example 4. It is a block diagram of the monitoring recorder of Example 4. It is a figure explaining the threshold value which determines the abnormal degree of Example 4. FIG. It is a figure explaining the threshold value which determines the abnormal degree of Example 4. FIG. It is a block diagram of the monitoring system of Example 5.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of a monitoring system according to an embodiment of the present invention. 101 is a camera, 102 is a threshold generation unit, 103 is a personal authentication unit, 104 is an abnormal behavior recognition unit, 105 is an abnormal degree comparison unit, 106 is an abnormal behavior notification unit, 107 is a monitoring device including 102 to 106, and 108 is a network It is. First, the flow of the present monitoring system will be described with reference to FIGS.
In FIG. 1, a video signal is input from a camera 101 to a monitoring device 107 via a transmission path indicated by a network 108.

  The monitoring device 107 authenticates a specific individual by the personal authentication unit 103 from the input video signal, and recognizes the abnormal behavior for each specific individual by the abnormal behavior recognition unit 104. That is, the personal authentication unit 103 detects human eyes, nose and mouth from the video signal supplied from the abnormal behavior recognition unit 104 based on the color and shape, for example, and authenticates a specific individual from their size and balance. To do. Furthermore, the abnormal behavior recognition unit 104 recognizes abnormal behavior that is inappropriate for a place such as staying at the same position for a long time, staring at one point, or being exposed to each specific individual, and obtains the degree of abnormal behavior. Further, the threshold generation unit 102 determines the threshold level of the degree of behavioral abnormality in response to a request from the abnormal behavior recognition unit 104.

  The abnormality degree of the specific individual determined by the abnormality recognition unit 104 and the threshold value of the abnormality degree determined by the threshold generation unit 102 are compared by the abnormality degree comparison unit 105 by the method described in FIG. In this case, the abnormality notification unit 106 notifies the abnormality monitoring device (not shown) monitored by the supervisor via the network 108.

  FIG. 2 is a diagram for explaining the threshold value for determining the degree of abnormality according to the first embodiment. The horizontal axis indicates time, and the vertical axis indicates the degree of abnormality based on abnormality determination obtained by the abnormal action recognition unit 104. FIG. 2 shows the relationship between the behavior of a specific person obtained from the video and the threshold value for determining the abnormal behavior. The plotted circles indicate the degree of behavioral abnormality obtained from an image at a certain time, and the white circles indicate the degree of behavioral abnormality determined to be equal to or less than the threshold A. A black circle indicates the degree of abnormality of the action determined to be greater than or equal to the threshold A. The vertical axis indicates a state in which the degree of abnormality increases as it goes up, and a black circle indicates an action detected as an abnormal action.

  For example, in FIG. 2, threshold A is provided as a threshold for determining an abnormality. As a method for determining this threshold, a method of calculating using an abnormal action image taken in the past or “Setting” in Patent Document 1 can be used. The threshold to be calculated can be obtained from a method in which an abnormality determination threshold is calculated and output using image data obtained by imaging a simulated intruder or the like.

  Here, when the threshold value A as shown in FIG. 2 and the abnormal degree of behavior as shown in plots 201 to 209 are obtained, the behavior that is determined as abnormal behavior in the conventional control is only the plot 209, and the plots 201 to 204 are shown. The plots 205 to 208 are not determined as abnormal behavior. In particular, the plots 205 to 208 are not determined to be abnormal behaviors even though the difference from the threshold A is very small. When the threshold setting is the threshold value A ′ in FIG. 2, all the plots 201 to 209 including the plot 209 are abnormal behavior even though the difference between the plot 209 and the threshold value A ′ is very small. It will not be judged. For this reason, it may not be possible to detect an action that is a precursor to an abnormal action such as an action immediately before causing the abnormal action.

  FIG. 3 is a diagram different from FIG. 2 for explaining the threshold value for determining the degree of abnormality in the first embodiment. Therefore, in this embodiment, as shown in FIG. 3, a threshold value A <b> 2 is provided, and the degree of abnormality of the action corresponding to the threshold value A to the threshold value A <b> 2 in a certain range from the threshold value A is determined. At this time, when the abnormal behavior indicated by the plots 305 to 308 determined in the range of the threshold A to the threshold A2 is due to a specific person, the threshold A is changed to the threshold A2. This threshold value change determination is triggered by the fact that the abnormal behavior in the range of threshold values A to A2 is determined a certain number of times, for example, three times as shown in FIG. By determining the abnormal behavior of the specific person as described above, the threshold A is changed to the threshold A2 in FIG. 3 for the plots 205 to 208 that were not determined to be abnormal behavior before the threshold change shown in FIG. Thereafter, the plot 308 is determined as abnormal behavior.

  FIG. 4 is a diagram different from FIGS. 2 and 3 for explaining the threshold for determining the degree of abnormality in the first embodiment. Here, the reason why the determination of abnormal behavior is a specific person will be described with reference to FIG. In FIG. 4, for example, it is assumed that four persons A to D (which is an example and generally may be an unspecified number) perform normal actions that are not abnormal actions. However, assuming that the normal behavior is determined to be abnormal behavior in very rare cases, the behavior of the person A happens to be a barely abnormal plot 401, and similarly, the behavior of the person B to the person D is also the threshold A and the threshold A2. It is assumed that the plots 402 to 404, which are the behaviors of the abnormal behavior just between, are obtained. Then, the abnormal behavior corresponding to the threshold value A to the threshold value A2 is first determined continuously three times in the plots 401 to 403.

  Therefore, when a determination is made without specifying a person, the determination threshold A for abnormal behavior is changed to a threshold A2, and thus the plot 404 that is a normal action of the person D is erroneously determined to be abnormal behavior. End up. Therefore, in the present embodiment, improvement is further made with respect to FIG. 4, and a specific individual is authenticated by the personal authentication unit 103 of FIG. 1, and then abnormal behavior is determined for each individual. For example, the determination threshold value for a specific person A is changed to the threshold value A2 when the person A performs an action that is between the threshold value A and the threshold value A2 three times. By identifying a person and determining abnormal behavior, the normal behavior as described above is determined to be abnormal behavior, or behavior that is close to abnormal behavior when consciously modest behavior is performed so that it is not determined to be abnormal behavior This makes it possible to accurately determine abnormal behavior.

3 and 4, when the threshold value generation unit 102 changes the threshold value from A to A2, it is not limited to the number of times determined as abnormal behavior, but the period (time) during which abnormal behavior is continued. You may make it change according to.
Moreover, about the determination of abnormality, the example shown using the image | video imaged with the monitoring apparatus was shown. However, the present invention is not limited to this, and the abnormality may be determined by detecting sounds, voices, infrared rays, pressures, vibrations, and the like generated by the captured person with a sensor.

Next, as a second embodiment, a case where a threshold value for determining abnormal behavior is changed according to a time zone will be described.
FIG. 5 is a diagram illustrating threshold values for determining the degree of abnormality in the second embodiment. In FIG. 5, the time from 0:00 to 9:00 is the threshold value B, and the time from 9:00 to 24:00 is the threshold value A. That is, the threshold for abnormal behavior is set low in the midnight from 0:00 to 9:00 and the daytime threshold from 9:00 to 24:00 is set high. In this example, it is assumed that when an abnormal action is performed in the midnight, the action is conspicuous so that the abnormal action person is careful so that his / her action does not become an abnormal action.

  FIG. 6 is a diagram different from FIG. 5 for explaining the threshold value for determining the degree of abnormality in the second embodiment. In this case, the threshold A2 is provided for the threshold A as in FIG. The degree of abnormality of the action corresponding to the threshold value A to the threshold value A2 and the threshold value B to the threshold value B2 of the region in the certain range is determined. When the abnormal behavior determined in the range of the threshold A to the threshold A2 and the threshold B to the threshold B2, that is, the abnormal behavior of the plots 501 to 503 and 505 to 507 in FIG. 5 is caused by a specific person, the threshold B is set to the threshold B2. Further, the threshold A is changed to the threshold A2. The determination of the former threshold change is triggered by the fact that abnormal behavior in the range of threshold B to threshold B2 is determined a certain number of times as in FIG. 2, and in the example of FIG. The determination of the latter threshold change is triggered by the fact that the abnormal behavior in the range of the threshold A to the threshold A2 is determined a certain number of times, in the example of FIG. As described above, the plots 501 to 508 that are not determined to be abnormal behavior in FIG. 5 are changed from the threshold value B to the threshold value B2 and the threshold value A to the threshold value A2, whereby the plots 604 and 608 in FIG. Can be determined.

Next, a third embodiment will be described with reference to FIG.
FIG. 7 is a block diagram of a monitoring system according to the third embodiment. 701 is a lens, 702 is an image sensor, 703 is an A / D conversion unit, 704 is an image processing unit, 705 is a Codec unit, 706 is an external I / F unit, and 707 is an abnormality detection unit. The abnormality detection unit 707 includes 102 to 106 in FIG.

  First, the flow of the imaging apparatus will be described with reference to FIG. Light incident through the lens 701 is photoelectrically converted by the image sensor 702 and converted into a digital signal by the A / D converter 703. The data converted into the digital signal is input to the abnormality detection unit 707 via the image processing unit 704. The abnormality detection unit 707 detects the degree of abnormality from the input video, and uses the detection result to control the image processing unit 704 and the Codec unit 705 as follows for the video determined to be abnormal behavior for a specific person. To process. That is, the image processing unit 704 performs processing for increasing the size of the recorded video by increasing the size of the recorded video and increasing the bit rate of the video so that the person can be easily identified. And the Codec unit 705 is controlled. The processed video data is output from the external I / F unit 706 to the outside of the imaging device, and transmitted to a recording device (monitoring recorder), a monitoring device (not shown), etc. via a route such as a network. The

  The abnormality detection method in the abnormality detection unit 707 may be the same as that in the first and second embodiments, but the third embodiment is different in that the abnormality detection unit 707 is included in the monitoring camera. Normally, all images shot by a camera are transferred to a monitoring device such as a surveillance recorder. However, as the quality of surveillance camera images continues to increase, it is required to efficiently use the bandwidth of the transmission path. However, only the video when the abnormality occurs is transferred, or the video with improved image quality of the portion where the abnormality has occurred is transferred. According to the third embodiment, since the abnormality detection unit 707 is included in the imaging device, not only a video at the time of abnormality but also a video that is a sign of abnormal behavior can be recorded on the recorder via the network. Therefore, according to the present embodiment, the detection omission related to the abnormal behavior is reduced, so that the situation where the video at the time of abnormality is not transferred and recorded in the monitoring recorder can be prevented. Although an example in which the recorder is connected to the network has been described here as a specific example, a configuration in which the recorder is built in the monitoring system may be employed.

Next, a fourth embodiment will be described with reference to FIG. 8, FIG. 9, and FIG.
FIG. 8 is a block diagram of the monitoring system of the fourth embodiment. 801 is a surveillance camera, 802 is a network, 803 is a surveillance recorder, 804 is a storage, 805 is another surveillance system connected to the network, 806 is a current video, 807 is a surveillance system acquired from another area, for example It is a past video.

  FIG. 9 is a block diagram of the monitoring recorder 803 according to the fourth embodiment. Reference numeral 901 denotes a threshold generation unit, 902 a personal authentication unit, 903 an abnormal behavior recognition unit, 904 an abnormality degree comparison unit, 905 a database search unit (hereinafter referred to as a DB search unit), and 906 an abnormality notification unit. 9 may have the same configuration as 102 to 106 in FIG.

  The video captured by the monitoring camera 801 is transmitted to the monitoring recorder 803 via the network 802, and the abnormal behavior is determined by the monitoring recorder 803. Depending on the determination, a specific person is searched from past images, for example. In this case, in the fourth embodiment, not only information from the monitoring camera 801 as shown by 806, for example, but also a monitoring system 805 for other areas, etc. By acquiring the video 807 of the corresponding person from the above, it is possible to search for abnormal behavior from a wide range of video data.

Next, a method for determining abnormal behavior of the monitoring recorder 803 will be described with reference to FIG.
In the monitoring recorder 803, a specific person is authenticated by the personal authentication unit 902, and abnormal behavior for each specific person is detected by the abnormal action recognition unit 903, as in 101 to 103 of the first embodiment. Then, the degree of abnormality of the behavior of the specific person is compared with the threshold levels A and A2 of the degree of abnormality determined by the threshold generation unit 901 by the degree of abnormality comparison unit 904. FIG. 10 shows the relationship between the threshold value at that time and the degree of abnormality of a specific individual.

FIG. 10 is a diagram illustrating threshold values for determining the degree of abnormality in the fourth embodiment.
In the fourth embodiment, as shown in FIG. 10, when the abnormal action recognition result plot 1001 determined from the current video is between the threshold value A and the threshold value A2, the specific person recognized as the abnormal action in the plot 1001 is stored in the DB. The search unit 905 searches for, for example, past video information 807 acquired from a monitoring system in another area.
FIG. 11 shows an example of the relationship between the threshold value of the person in the plot 1001 in which past video information has been found and the degree of abnormality.

FIG. 11 is a diagram different from FIG. 10 for explaining the threshold value for determining the degree of abnormality in the fourth embodiment.
In the example of FIG. 11, since there is no abnormal behavior continuously corresponding to the thresholds A to A2 in the same monitoring area, it is not determined as abnormal behavior in the first embodiment, but in the fourth embodiment, the past of other areas is determined. Abnormal behavior recognition unit 903 determines an abnormality from the image. For this reason, abnormal behaviors corresponding to threshold A to threshold A2 are detected as plots 1101 and 1103 from other monitoring areas 1 and as plots 1104 and 1106 from other monitoring areas 2, so the frequency of these abnormal behaviors is detected. Accordingly, the action of the person in the plot 1001 is determined as an abnormal action. In the present embodiment, data acquired from other areas is described as video, but may be metadata generated at the time of abnormality recognition.

In the above description, the DB search unit 905 of the monitoring recorder 803 searches for a specific person from the past video information 807 acquired from the monitoring system in another area. However, the present embodiment is not limited to this.
For example, current video information acquired from another monitoring system may be searched. If other monitoring systems are installed in close proximity, the same person may be photographed. In this case, 905 may be called a video search unit.
Moreover, you may search the video information which the monitoring system itself acquired in the past. In this case, the DB search unit 905 may search the storage 804 for video information acquired in the past.

Next, a fifth embodiment will be described with reference to FIG.
FIG. 12 is a block diagram of a monitoring system according to the fifth embodiment. Reference numeral 1201 denotes a lens, 1202 denotes an image sensor, 1203 denotes an A / D conversion unit, 1204 denotes an image processing unit, 1205 denotes a Codec unit, 1206 denotes an external I / F unit, 1207 denotes an abnormality detection unit, and 1208 denotes a network. The abnormality detection unit 1207 includes the same components as 901 to 905 and the abnormality detection unit 906 in FIG. The monitoring camera 1210 and the monitoring camera 1211 connected via the network 1208 may have the same configuration as the monitoring camera 1209 configured by the above 1201 to 1207.

The abnormality detection of the abnormality detection unit 1207 of the fifth embodiment may be the same as that of the first and second embodiments, and the monitoring camera includes the abnormality detection unit as in the third embodiment.
The flow of the imaging apparatus is the same as in the third embodiment and FIG. 7, and the captured video is input to the abnormality detection unit 1207 via the image processing unit 1204. The abnormality detection of the abnormality detection unit 1207 is the same as in the first and second embodiments. However, with respect to the fifth embodiment, the abnormality detection unit 1207 is included in the monitoring camera, and not only a video at the time of abnormality but also a sign of abnormal behavior. Can be recorded on the recorder via the network. Therefore, according to the present embodiment, since the omission of detection related to abnormal behavior is reduced, the situation where the video at the time of abnormality is not transferred and recorded on the monitoring recorder is prevented.

  Furthermore, in the fifth embodiment, the data used for abnormality detection by the monitoring camera 1209 is not only the video imaged by the monitoring camera 1209, but is also captured by other monitoring cameras such as the monitoring cameras 1210 and 1211 via the network, or being captured. For example, it is possible to acquire an image from another monitoring camera or recorder connected to the network via the network and perform abnormality detection. In this embodiment, data acquired from other surveillance cameras is video, but it may be metadata generated at the time of abnormality recognition.

  The embodiment described so far is merely an example and does not limit the present invention. For example, other embodiments such as a block configuration of a monitoring camera or a monitoring system and a method for setting a threshold value for determining an abnormality can be considered, but all fall within the scope of the present invention.

DESCRIPTION OF SYMBOLS 101: Camera, 102: Threshold determination part, 103: Personal authentication part, 104: Abnormal action recognition part, 105: Abnormality comparison part, 106: Abnormal action notification part, 107: Monitoring apparatus comprised by 102-106, 108 : Network, 701: Lens, 702: Image sensor, 703: A / D conversion unit, 704: Image processing unit, 705: Codec unit, 706: External I / F unit, 707: Abnormality detection unit, 801: Surveillance camera, 802: Network, 803: Monitoring recorder, 804: Storage, 805: Other monitoring system connected to the network, 806: Current video, 807: Past video of the monitoring system acquired from other area, 901: Threshold generation 902: Personal authentication unit 903: Abnormal behavior recognition unit 904: Abnormality comparison unit 905: Database search unit 906: Different Notification unit, 1201: Lens, 1202: Image sensor, 1203: A / D conversion unit, 1204: Image processing unit, 1205: Codec unit, 1206: External I / F unit, 1207: Abnormality detection unit, 1208: Network, 1210 : Surveillance camera, 1211: surveillance camera.

Claims (7)

A monitoring device that monitors abnormal behavior of a subject,
A personal authentication unit that is supplied with an image of the subject and identifies a person included in the image;
An abnormal action recognizing unit that is supplied with an image of the subject and detects an abnormal degree of the behavior of the person included in the image;
A threshold generation unit that determines a threshold of the degree of abnormality of the behavior according to the person specified by the personal authentication unit;
An abnormality degree comparison unit that compares the degree of abnormality of the behavior detected by the abnormal action recognition unit with the threshold value set by the threshold value generation unit and determines the presence or absence of the abnormal behavior of the person;
When the abnormality degree comparison unit determines that there is an abnormal behavior of the person, it has an abnormality notification unit for notifying an external device of the abnormality,
When the abnormality degree of the behavior detected by the abnormal behavior recognition unit is smaller than the threshold value generated by the threshold value generation unit within a predetermined range, and / or continues for a predetermined period If you do
The threshold generation unit reduces the threshold by a predetermined amount. The monitoring device according to claim 1,
Another monitoring device has a video search unit that searches a video of a person being imaged via a network,
When the abnormal degree of the behavior detected by the abnormal behavior recognition unit is smaller than the threshold generated by the threshold generation unit within a predetermined range,
The personal authentication unit refers to a video of a person being imaged by another monitoring device searched by the video search unit,
If the video of the person is searched by the video search unit,
The threshold generation unit reduces the threshold by a predetermined amount. A monitoring system that monitors abnormal behavior of a subject,
An imaging unit for imaging the subject and converting it into a digital signal;
An image processing unit for converting a bit rate of the digital signal supplied from the imaging unit;
An external interface unit for sending a digital signal after bit rate conversion supplied from the image processing unit to the outside of the monitoring system;
An abnormality detection unit for controlling a bit rate conversion operation in the image processing unit;
The abnormality detection unit
A personal authentication unit for supplying a video of the subject imaged by the imaging unit and identifying a person included in the video;
An abnormal behavior recognizing unit that is supplied with a video of the subject imaged by the imaging unit and detects an abnormal degree of the behavior of the person included in the video;
A threshold generation unit that determines a threshold of the degree of abnormality of the behavior according to the person specified by the personal authentication unit;
An abnormality degree comparison unit that compares the degree of abnormality of the behavior detected by the abnormal behavior recognition unit with the threshold value determined by the threshold value generation unit and determines the presence or absence of the abnormal behavior of the person;
When the abnormality degree of the behavior detected by the abnormal behavior recognition unit is smaller than the threshold value generated by the threshold value generation unit within a predetermined range, and / or continues for a predetermined period In this case, the threshold generation unit reduces the threshold by a predetermined amount,
And controlling the image processing unit to increase the bit rate of the digital signal after bit rate conversion in the image processing unit when the abnormality level comparing unit determines that the person has an abnormal behavior. A monitoring system characterized by The monitoring system according to claim 3,
The abnormality detection unit includes a video search unit that searches for video of a person being imaged by another monitoring device via the external interface unit,
When the abnormal degree of the behavior detected by the abnormal behavior recognition unit is smaller than the threshold generated by the threshold generation unit within a predetermined range,
The personal authentication unit refers to a video of a person being imaged by another surveillance system searched by the video search unit,
If the video of the person is searched by the video search unit,
The threshold generation unit reduces the threshold by a predetermined amount. The monitoring system according to claim 3,
The abnormality detection unit includes a database search unit that searches for images of a person captured in the past by another monitoring system via the external interface unit,
When the abnormal degree of the behavior detected by the abnormal behavior recognition unit is smaller than the threshold generated by the threshold generation unit within a predetermined range,
The personal authentication unit refers to an image of a person captured in the past by another monitoring system searched by the database search unit,
If the video of the person is searched by the database search unit,
The threshold generation unit reduces the threshold by a predetermined amount. The monitoring device according to claim 1,
It has a storage unit that stores videos of people supplied in the past,
When the abnormal degree of the behavior detected by the abnormal behavior recognition unit is smaller than the threshold generated by the threshold generation unit within a predetermined range,
The personal authentication unit refers to a video of a person supplied in the past stored in the storage unit,
If the video of the person is stored in the storage unit,
The threshold generation unit reduces the threshold by a predetermined amount. The monitoring device according to claim 1,
Another monitoring device has a database search unit that refers to a video of a person supplied in the past,
When the abnormal degree of the behavior detected by the abnormal behavior recognition unit is smaller than the threshold generated by the threshold generation unit within a predetermined range,
The personal authentication unit refers to a video of a person previously supplied by another monitoring device searched by the database search unit,
If the video of the person is searched by the database search unit,
The threshold generation unit reduces the threshold by a predetermined amount.

Images