System and method for dynamic response to emergency alerts based on situational awareness

Abstract

The video stream captures at least a portion of an area of the facility, and performs video analysis to identify one or more video-identified events. One or more sensor-recognized events are received, and a current state of the area of the facility is classified as one of a plurality of predetermined state categories based at least in part on the one or more video-recognized events and the one or more sensor-recognized events. The activation of the emergency button causes an emergency alert to be issued, where the emergency alert is issued as a silent emergency alert when the current state of the area is classified in a first predetermined state category, and the emergency alert is issued as a sound emergency alert when the current state of the area is classified in a second predetermined state category.

Description

Technical Field

[0001] The present disclosure generally relates to security systems, and more particularly to security systems with emergency alerts. Background Art

[0002] Commercial security systems include intrusion-related detection and life-safety-related detection, such as fire and carbon monoxide. Commercial security systems can include emergency alerts that an individual may trigger in an emergency. Commercial security systems can trigger an audible alert, where a signal is sent to the authorities and a local siren is sounded. Commercial security systems can trigger a silent alert, where a signal is sent to the authorities, but the local siren does not sound.

[0003] In some cases, depending on the specific situation, it is best for a commercial security system to trigger an audible alert. A medical emergency is an example of a situation where an audible alert may be the best choice. In some cases, depending on the specific situation, it is best for a commercial security system to trigger a silent alert. An individual being threatened with a firearm is an example of a situation where a silent alert may be the best choice. A commercial security system can be preconfigured to respond in a specific way, but this preconfiguration is static and does not change dynamically, depending on the current situation. What is desired is a method and system such that a commercial security system can determine the current situation in which an emergency alert has been triggered and respond appropriately. Summary of the Invention

[0004] The present disclosure generally relates to security systems, and more particularly to security systems with emergency alerts. An example can be found in a method for responding to activation of an emergency button in an area of a facility. An exemplary method includes receiving a video stream that captures at least a portion of an area of a facility and performing video analysis on the video stream to identify one or more video-identified events associated with the area of the facility. One or more sensor-identified events sensed by one or more sensors associated with the area of the facility are also received. The current state of the area of the facility is classified as one of a plurality of predetermined state categories at least in part based on the one or more video-identified events and the one or more sensor-identified events. A notification that an emergency button has been activated by a user in the area of the facility is received. In response to receiving the notification that the emergency button has been activated, an emergency alert associated with the area of the facility is issued, where the emergency alert is issued as a silent emergency alert when the current state of the area is classified as a first predetermined state category of the plurality of predetermined state categories (e.g., an individual is being coerced), and the emergency alert is issued as an audible emergency alert when the current state of the area is classified as a second predetermined state category of the plurality of predetermined state categories (e.g., an individual is suffering from a medical condition such as a fall).

[0005] Another example may exist in a system. The exemplary system includes an emergency button, a video camera, one or more sensors, an output, and a controller. The emergency button is associated with an area of a facility. The video camera is configured to capture a video stream of at least a portion of the area of the facility. The one or more sensors are associated with the area of the facility to sense one or more sensor-identified events associated with the area of the facility. The output is configured to issue an alert. The controller is operatively coupled to the emergency button, the video camera, the one or more sensors, and the output. The controller is configured to perform video analysis on the video stream to identify one or more video-identified events associated with the area of the facility, and classify a current state of the area of the facility as one of a plurality of predetermined state categories at least in part based on the one or more video-identified events and the one or more sensor-identified events. The controller is configured to receive a notification that the emergency button has been activated, and in response to receiving the notification that the emergency button has been activated, issue an emergency alert associated with the area of the facility, wherein when the current state of the area is classified as a first predetermined state category of the plurality of predetermined state categories, the emergency alert is issued as a silent emergency alert, and when the current state of the area is classified as a second predetermined state category of the plurality of predetermined state categories, the emergency alert is issued as an audible emergency alert.

[0006] Another example may exist in a non-transitory computer-readable medium storing instructions. When the instructions are executed by one or more processors, the one or more processors are caused to receive a video stream capturing at least a portion of an area of a facility and perform video analysis on the video stream to identify one or more video-identified events associated with the area of the facility. The one or more processors are caused to receive one or more sensor-identified events sensed by one or more sensors associated with the area of the facility, and classify a current state of the area of the facility as one of a plurality of predetermined state categories at least in part based on the one or more video-identified events and the one or more sensor-identified events. The one or more processors are caused to receive a notification that the emergency button has been activated, and in response to receiving the notification that the emergency button has been activated, issue an emergency alert associated with the area of the facility, wherein when the current state of the area is classified as a first predetermined state category of the plurality of predetermined state categories, the emergency alert is issued as a silent emergency alert, and when the current state of the area is classified as a second predetermined state category of the plurality of predetermined state categories, the emergency alert is issued as an audible emergency alert.

[0007] The foregoing Summary is provided to facilitate understanding of some innovative features unique to the present disclosure and is not intended as a complete description. A full understanding of the present disclosure can be obtained by considering the entire specification, claims, drawings, and abstract as a whole. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The present disclosure can be more fully understood in connection with the following description of various examples, considered in conjunction with the accompanying drawings, in which:

[0009] Figure 1 is a schematic block diagram showing an exemplary security system;

[0010] Figure 2 is a schematic block diagram showing an exemplary security system; and

[0011] Figure 3 is a flowchart showing an exemplary method for responding to activation of an emergency button.

[0012] While the present disclosure is subject to various modifications and alternative forms, specific details thereof have been shown by way of example in the drawings and will be described in detail. It should be understood, however, that the intention is not to limit the present disclosure to the particular examples described. On the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the present disclosure. DETAILED DESCRIPTION

[0013] The following description should be read with reference to the drawings, in which like elements in different drawings are numbered in the same manner. The drawings are not necessarily to scale and depict examples that are not intended to limit the scope of the present disclosure. While examples of various elements are illustrated, those skilled in the art will recognize that many of the examples provided have suitable alternatives that can be utilized.

[0014] It is assumed herein that all numbers are modified by the term "about" unless the context clearly dictates otherwise. The recitation of numerical ranges by endpoints includes all numbers subsumed within that range (e.g., 1 to 5 includes 1, 1.5, 2, 2.75, 3, 3.80, 4, and 5).

[0015] As used in this specification and the appended claims, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. As used in this specification and the appended claims, the term "or" is generally employed in its sense including "and / or" unless the context clearly dictates otherwise.

[0016] It should be noted that when "an embodiment", "some embodiments", "other embodiments", etc. are mentioned in the specification, it indicates that the described embodiments may include specific features, structures or characteristics, but each embodiment does not necessarily include the specific feature, structure or characteristic. Moreover, these phrases do not necessarily refer to the same embodiment. Additionally, when a specific feature, structure or characteristic is described in connection with an embodiment, it is contemplated that the feature, structure or characteristic is described in connection with an embodiment, and it is contemplated that the feature, structure or characteristic can be applied to other embodiments whether or not explicitly described, unless there is an express contrary statement.

[0017] Figure 1 is a schematic block diagram showing an exemplary security system 10. The exemplary security system 10 includes an emergency button 12 associated with an area of a facility. The emergency button 12 can be configured to enable a user to press or otherwise activate the emergency button 12 when the user sees, hears, or otherwise detects or suspects a potential problem. For example, this can include a suspected fire or visible smoke. This can include a sudden fight. This can include a medical emergency. The emergency button 12 can be a wired emergency button or a wireless emergency button. The exemplary security system 10 includes a video camera 14 that is configured to capture a video stream of at least a portion of the area of the facility, such as including a portion of the facility corresponding to the location of the emergency button 12. In some cases, the exemplary security system 10 includes a plurality of video cameras that are configured to capture a plurality of video streams of at least a portion of the area of the facility. The exemplary security system 10 includes a plurality of sensors 16 that are respectively labeled 16a, 16b to 16n. The exemplary security system 10 can include any number of sensors 16, including intrusion sensors such as motion sensors, glass break sensors, security panel keypads, and tamper sensors. The sensors 16 can include one or more access control sensors such as card readers or biometric readers. The sensors 16 can include one or more gas sensors, one or more temperature sensors, one or more humidity sensors, and / or one or more fire sensors. These are just examples. The exemplary security system 10 includes an output 18 that is configured to issue an alarm. The output 18 can include a sound device such as a siren or a speaker. The output 18 can include a communication channel that enables a warning or an alarm to be sent to, for example, an authority or security personnel.

[0018] Exemplary security system 10 includes a controller 20 that is operatively coupled to an emergency button 12, one or more video cameras 14, one or more sensors 16, and an output 18. The controller 20 is configured to perform video analysis on a video stream from the one or more video cameras 14 to identify one or more video-identified events associated with an area of a facility. The controller 20 is configured to classify a current state of the area of the facility as one of a plurality of predetermined state categories at least in part based on the one or more video-identified events and one or more sensor-identified events. The controller 20 is configured to receive a notification that the emergency button has been activated and, in response, emit an emergency alert associated with the area of the facility via the output, wherein when the current state of the area is classified as a first predetermined state category of the plurality of predetermined state categories, the emergency alert is emitted as a silent emergency alert, and when the current state of the area is classified as a second predetermined state category of the plurality of predetermined state categories, the emergency alert is emitted as an audible emergency alert. In some cases, the controller 20 can be configured to fuse video analysis (e.g., video-identified events), intrusion sensor data (e.g., intrusion sensor-identified events), and access control sensor data (e.g., access control sensor events) when classifying the current state of the area as the first predetermined state category of the plurality of predetermined state categories or the second predetermined state category of the plurality of predetermined state categories.

[0019] In some cases, the controller 20 can be a distributed controller, which means that a portion of the controller 20 can be co-located with the video cameras 14. In some cases, the portion of the controller 20 that is co-located with the video cameras 14 can be configured to perform at least some of the video analysis performed on the video stream of the video cameras 14 to identify one or more video-identified events associated with an area of a facility. In some cases, a portion of the controller 20 can be implemented on a server in a cloud environment. In some cases, the controller 20 can implement an artificial intelligence and / or machine learning (AI / ML) engine that is trained to identify one or more video-identified events associated with an area of a facility. In some cases, the controller 20 can include an artificial intelligence and / or machine learning (AI / ML) engine that is trained to identify the current state of the area of the facility and classify the current state of the area of the facility as one of a plurality of predetermined state categories at least in part based on the one or more video-identified events and one or more sensor-identified events.

[0020] Figure 2is a schematic block diagram showing an exemplary security system 22, which can be considered an example of security system 10( Figure 1 ). The video module 24 receives video streams from a plurality of video cameras 26 respectively labeled 26a through 26n. The access module 28 communicates with a plurality of readers 30 respectively labeled 30a through 30n. For example, the access module 28 can communicate with the readers 30 to indicate whether a particular user presenting their credentials (such as an access card or biometric fingerprint) is authorized to proceed. The intrusion module 32 communicates with a plurality of sensors 34 respectively labeled 34 through 34n. The sensors 34 can include any intrusion sensors, such as motion sensors, glass break sensors, security panel keypads, and tamper sensors. In some cases, the sensors 34 can include one or more access control sensors, such as card readers or biometric readers. The sensors 34 can include one or more gas sensors, one or more temperature sensors, one or more humidity sensors, and / or one or more fire sensors. These are just examples.

[0021] The video module 24, the access module 28, and the intrusion module 32 can each communicate with a control panel 36. In some cases, the control panel 36 can be considered to be Figure 1An example of the controller 20 shown. The control panel 36 implements control logic that includes a security application module 38, a data pool 40, a data analysis block 42, and a machine learning block 44. Events and / or alerts detected by the video module 24, the access and egress module 28, and the intrusion module 32 are fed into the security application module 38, and then the security application module sends the events and alerts to the data pool 40. The data pool 40 filters the events and / or alerts and sends the filtered events and / or alerts to the data analysis block 42. The data pool 40 can filter the events and / or alerts, for example, by removing outliers, duplicates, and / or one or more event or alert types of no interest. In the example shown, the data analysis block 42 analyzes the filtered events and / or alerts and provides the fused events and / or alerts to the machine learning block 44. The events and / or alerts can be related to each other. For example, the detection of an individual in an area by a motion sensor and the detection of the individual in the area by a video analysis algorithm can correlate or fuse these events. When the video analysis algorithm includes a facial recognition algorithm, the uniquely identified individual can be recognized and the individual can be associated or fused with the motion detection event detected by the motion sensor. In another example, the access and egress module 28 can report the entry of uniquely identified individuals into an area (e.g., via a card swipe or biometric scan), and these uniquely identified individuals can be associated or fused with the people detected in the area by the video analysis algorithm. These are just examples. The machine learning block 44 can include an artificial intelligence and / or machine learning (AI / ML) engine that is trained to recognize the current state of an area of a facility and classify the current state of the area of the facility into one of a plurality of predetermined state categories based at least in part on one or more video-identified events and one or more sensor-identified events. In the example shown, various measures are passed back and forth between the security application module 38 and the machine learning block 44. For example, the security application module 38 can issue a warning based on the current state of the area determined by the machine learning block 44.

[0022] In some cases, the control panel 36 may receive multiple events or warnings, including some events that may warrant an audible alarm and some events that may warrant a silent alarm, and may need to weight various conflicting events to determine whether to issue an audible or silent alarm. In some cases, an alarm may belong to one of three different categories. "Normal status" means that although one or more cameras 26, one or more readers 30, or one or more sensors 34 may indicate an alarm situation, no one has initiated an emergency alarm. In the normal state, for whether to trigger an audible or silent alarm for a given alarm condition, the control panel 36 may rely on its initial pre-configuration. Some alarms may belong to the intrusion emergency category and may be pre-configured to issue a silent alarm. Some alarms may belong to the fire or medical emergency or fire emergency category and may be pre-configured to issue an audible alarm. The following table provides some examples of possible alarms and alarm categories:

[0023] Table I

[0024] In some cases, the initial pre-configuration settings may be overridden based on the situational context of the corresponding area, i.e., whether to trigger an audible or silent alarm in response to pressing the emergency button for a given alarm condition. For example, in some cases, multiple events and / or alarms may be triggered during a common time frame. For example, assume there is a detected tailgating in and out event and a flame detection warning. The tailgating in and out event may have a medium weight (and a silent alarm), and the flame detection warning may have a high weight (and an audible alarm). In this scenario, the system may generate an audible alarm in response to pressing the emergency button because the weight of the flame detection warning is higher than that of the tailgating in and out event. In some cases, there may be conflicting warnings. For example, assume there is an alarm for identifying a weapon / gun (high weight and silent alarm) and a flame detection warning (high weight and audible alarm). In this example, the system may determine that the weight of the flame detection warning is higher than that of the warning for identifying a gun because a fire may cause greater damage and loss of life. These are just examples.

[0025] In some cases, the situational context of the corresponding area (e.g., normal, burglary emergency, medical emergency) can be identified by, for example, performing pattern analysis on one or more video-identified events associated with the area and / or one or more sensor-identified events associated with the area (e.g., intrusion, tampering, fire, smoke, entering a duress code when deactivating the security system, entering an area bypass when activating the security system). The pattern analysis can reveal phenomena and predict the situational context of the area, and can apply an audible or silent alarm in response to pressing the emergency button based on the situational context.

[0026] Figure 3is a flowchart showing an exemplary method 50 for responding to activation of an emergency button (such as emergency button 12) in an area of a facility. The exemplary method includes receiving a video stream that captures at least a portion of the area of the facility, as indicated at block 52. Performing video analysis on the video stream to identify one or more video-recognized events associated with the area of the facility, as indicated at block 54. In some cases, the one or more video-recognized events may include one or more of the recognized behavior of one or more persons and the presence of recognized objects. Receiving one or more sensor-recognized events sensed by one or more sensors associated with the area of the facility, as indicated at block 56. In some cases, the one or more sensors associated with the area of the facility may include one or more intrusion sensors. Examples of intrusion sensors include motion sensors, glass break sensors, security panel keypads, and tamper sensors. The one or more sensors associated with the area of the facility may include one or more access control sensors, one or more gas sensors, one or more temperature sensors, one or more humidity sensors, and one or more fire sensors. For example, access control sensors may include card readers and biometric readers. At least some of the one or more sensor-recognized events may include one or more alarms.

[0027] Classifying the current state of the area of the facility as one of a plurality of predetermined state categories based at least in part on the one or more video-recognized events and the one or more sensor-recognized events, as indicated at block 58. A first predetermined state category of the plurality of predetermined state categories may correspond to an intrusion emergency category, and a second predetermined state category of the plurality of predetermined state categories may correspond to a medical emergency category and / or a fire emergency category. Receiving a notification that an emergency button has been activated by a user in the area of the facility, as indicated at block 60. In response to receiving the notification that the emergency button has been activated, issuing an emergency alarm associated with the area of the facility, wherein when the current state of the area is classified as the first predetermined state category of the plurality of predetermined state categories, the emergency alarm is issued as a silent emergency alarm, and when the current state of the area is classified as the second predetermined state category of the plurality of predetermined state categories, the emergency alarm is issued as an audible emergency alarm, as indicated at block 62. In some cases, the silent emergency alarm may not activate any sirens associated with the area of the facility, and the audible emergency alarm may activate one or more sirens associated with the area of the facility.

[0028] In some cases, method 50 may further include providing one or more video-identified events and one or more sensor-identified events to an artificial intelligence and / or machine learning (AI / ML) engine, as indicated at block 64. In some cases, over time, the AI / ML engine may be trained to identify the current state of a region of a facility and classify the current state of the region of the facility as one of a plurality of predetermined state categories, at least in part, based on one or more video-identified events and one or more sensor-identified events, as indicated at block 66.

[0029] In some cases, over time, the AI / ML engine may be trained to identify the current state of a region of a facility and classify the current state of the region of the facility as one of a plurality of predetermined state categories, at least in part, based on a particular combination of current active events and / or alerts of a security system, such as those shown in Table I. In some cases, the AI / ML engine may be trained by providing a variety of different combinations of events and / or alerts of the security system during a training phase and correcting the AI / ML engine until the AI / ML engine correctly classifies the current state of the region of the facility with a desired accuracy. It is contemplated that the AI / ML engine may consider not only the current active events and / or alerts of the security system, but may also be trained using historical events and / or alerts of the security system to help achieve a more robust and accurate classification. In some cases, the AI / ML engine is continuously trained by having an operator of the security system verify the classification of the AI / ML engine and correcting the AI / ML engine as necessary. The correction results are used as input for retraining the AI / ML engine. In some cases, the AI / ML engine may include a support vector machine (SVM) learning algorithm, which may be particularly adept at classifying the current state of a region of a facility as one of a plurality of predetermined state categories. However, any other suitable learning algorithm may be used.

[0030] Although several illustrative embodiments of the present disclosure have been described as such, those skilled in the art will readily appreciate that other embodiments can be made and used within the scope of the appended claims. However, it should be understood that the present disclosure is illustrative in many respects. Changes may be made to the details, particularly to details related to the shape, size, arrangement of parts, and the exclusion and order of steps, without departing from the scope of the present disclosure. Of course, the scope of the present disclosure is defined in the language of the appended claims.

Claims

1. A method for responding to activation of a panic button in an area of ​​a facility, the method comprising: receiving a video stream capturing at least a portion of the area of ​​the facility; performing video analysis on the video stream to identify one or more video-identified events associated with the area of ​​the facility; receiving one or more sensor-identified events sensed by one or more sensors associated with the area of ​​the facility; classifying a current state of the area of ​​the facility into one of a plurality of predetermined state categories based at least in part on the one or more video identified events and the one or more sensor identified events; receiving notification of activation of a panic button by a user in said area of ​​said facility; as well as In response to receiving the notification that the panic button has been activated, a panic alarm associated with the area of ​​the facility is issued, wherein the emergency alarm is issued as a silent emergency alarm when the current state of the area is classified as a first predetermined status category among the multiple predetermined status categories, and the emergency alarm is issued as an audible emergency alarm when the current state of the area is classified as a second predetermined status category among the multiple predetermined status categories.

2. The method according to claim 1, wherein: The silent panic alarm does not activate any sirens associated with the area of ​​the facility, and the audible panic alarm activates one or more sirens associated with the area of ​​the facility.

3. The method according to claim 1, further comprising: Providing one or more of the video-recognized events and one or more sensor-recognized events to an artificial intelligence and / or machine learning (AI / ML) engine; as well as The AI / ML engine is trained over time to identify the current state of the area of ​​the facility and classify the current state of the area of ​​the facility into one of the plurality of predetermined state categories based at least in part on the one or more video-identified events and the one or more sensor-identified events.

4. The method according to claim 1, wherein: The one or more sensors associated with the area of ​​the facility include one or more intrusion sensors.

5. The method according to claim 4, wherein: The one or more intrusion sensors include one or more of a motion sensor, a glass break sensor, a security panel keypad, and a tamper sensor.

6. The method according to claim 1, wherein: The one or more sensors associated with the area of ​​the facility include one or more access control sensors, one or more gas sensors, one or more temperature sensors, one or more humidity sensors, and one or more fire sensors.

7. The method according to claim 6, wherein: The one or more sensors associated with the area of ​​the facility include one or more access control sensors, and wherein the one or more access control sensors include one or more of a card reader and a biometric reader.

8. The method according to claim 1, wherein: The one or more events recognized in the video include one or more of recognized behaviors of one or more persons and the presence of recognized objects.

9. The method according to claim 1, wherein: The first of the plurality of predetermined status categories corresponds to an intrusion emergency category, and the second of the plurality of predetermined status categories corresponds to a medical emergency category and / or a fire emergency category.

10. A system, comprising: a panic button associated with an area of ​​the facility; a video camera for capturing a video stream of at least a portion of the area of ​​the facility; one or more sensors associated with the area of ​​the facility for sensing one or more sensor-identified events associated with the area of ​​the facility; an output terminal, the output terminal being used to issue an alarm; a controller operatively coupled to the panic button, the video camera, the one or more sensors, and the output, the controller being configured to: performing video analysis on the video stream to identify one or more video-identified events associated with the area of ​​the facility; classifying a current state of the area of ​​the facility into one of a plurality of predetermined state categories based at least in part on the one or more video identified events and the one or more sensor identified events; receiving a notification that the panic button has been activated; as well as In response to receiving the notification that the panic button has been activated, a panic alarm associated with the area of ​​the facility is issued via the output terminal, wherein when the current state of the area is classified as a first predetermined status category among the multiple predetermined status categories, the emergency alarm is issued as a silent emergency alarm, and when the current state of the area is classified as a second predetermined status category among the multiple predetermined status categories, the emergency alarm is issued as an audible emergency alarm.

Images