VIDEO PATTERN RECOGNITION FOR AUTOMATING EMERGENCY SERVICE INCIDENT AWARENESS AND RESPONSE

• US 20100231714A1
• Filed: 03/12/2009
• Published: 09/16/2010
• Est. Priority Date: 03/12/2009
• Status: Active Grant

• Current Assignee: Generac Power Systems Incorporated (Generac Holdings Incorporated)
• Original Assignee: International Business Machines Corporation
• Inventors: KWAN, HUNG T., GIVENS, CHARLET N., MATHAI, SHIJU, FLORES, ROMELIA H.
• Granted Patent: US 8,368,754 B2
• Time in Patent Office: Days
• US Class Current: 348/143
• CPC Class Codes: G08B 13/19613 Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion
• CPC Class Codes: G08B 13/19695 Arrangements wherein non-video detectors start video recording or forwarding but do not generate an alarm themselves
• CPC Class Codes: G08B 25/006 Alarm destination chosen according to type of event, e.g. in case of fire phone the fire service, in case of medical emergency phone the ambulance
• CPC Class Codes: H04M 11/04 with alarm systems, e.g. fire, police or burglar alarm systems

Abstract

Video analysis capabilities can be integrated into an emergency response system and/or a video operations center (VOC) monitoring for emergency incidents. At least one computer program driven response action involving detecting patterns in video streams that comprise video of an incident can be performed. This response action can be directed to VOC personnel and/or to emergency response system personnel. Further, the response action can result in an automated responses to an emergency incident and/or to an alerting of humans monitoring video to focus on a particular video stream/incident. One such response action, which increases video capture activity at a geographic location of the incident, can occur responsive to a receipt of an emergency call by the emergency response system. Another response action can analyze at least one video stream of the geographic location and can automatically dispatch (in approximately real time) an emergency responder to the geographic location. Still another action can analyze the video stream of a geographic location, can determine a high likelihood of an emergency incident, and can provide an alert in a user interface of the emergency response system or a VOC. Video of the incident can also be provided in the user interface.

Claims

1. An emergency response method comprising:

integrating video analysis capabilities into at least one of an emergency response system and a video operations center (VOC) monitoring for emergency incidents; and

performing at least one computer program driven response action involving detecting patterns in video streams that comprise video of an incident, wherein the response action comprises at least one of;

(A) responsive to a receipt of an emergency call by the emergency response system for an incident occurring at a geographic location, automatically executing a programmatic action to increase video capture activity at the geographic location and detecting patterns in video streams acquired from the increased video capture activity;

(B) analyzing at least one video stream of a geographic location and automatically dispatching an emergency responder to the geographic location based upon results of analyzing the at least one video stream, wherein the analyzing and automatic dispatching occur in at least one of real-time and near real-time; and

(C) analyzing at least one video stream of a geographic location, determining a high likelihood of an emergency incident occurring at the geographic location and determining an incident type of the emergency incident based upon comparing patterns in the analyzed video stream against previously defined patterns associated with different incidents, automatically providing an alert in a user interface of at least one of the emergency response system and the video operations center (VOC), wherein said alert indicates the determined emergency incident, the determined incident type, and presents the video stream including a portion of video upon which the patterns were determined.

2. The method of claim 1, wherein said response action comprises:

responsive to a receipt of an emergency call by the emergency response system for an incident occurring at a geographic location, automatically executing a programmatic action to increase video capture activity at the geographic location and detecting patterns in video streams acquired from the increased video capture activity.

3. The method of claim 2, wherein the response action comprises:

detecting a plurality of video sources proximate to the geographic location;

routing streams from the video sources to at least one of the emergency response system and the video operations center (VOC);

executing pattern matching activities against each of the routed video streams;

conveying at least one of results from the executed pattern matching activities and the routed streams to at least one of emergency response personnel interfacing with the emergency response system, personnel of the video operations center (VOC), and to responders dispatched to the geographic location, wherein the detecting of video sources, the routing of streams, the executing of pattern matching activities, and the conveying of at least one of results and routed streams occurs in at least one of real-time and near real-time.

4. The method of claim 2, further comprising:

determining geospatial information from the emergency call, utilizing the geospatial information to determine the geographic location; and

executing the programmatic action to increase video capture activity that automatically obtains new video streams covering the geographic location that are routed to at least one of the emergency response system and the video operations center (VOC).

5. The method of claim 1, wherein said response action comprises:

analyzing at least one video stream of a geographic location and automatically dispatching an emergency responder to the geographic location based upon results of analyzing the at least one video stream, wherein the analyzing and automatic dispatching occur in at least one of real-time and near real-time.

6. The method of claim 5, further comprising:

aggregating results of analyzing the at least one video stream of the geographic location with data input into the emergency response system based upon real-time voice communications involving emergency response system personnel;

determining a likelihood of an emergency incident based upon the aggregated results of analyzing the video stream and the data input based upon real time voice communications; and

automatically dispatching the emergency responder only when the determined likelihood of the emergency incident exceeds a previously established threshold.

7. The method of claim 1, further comprising:

analyzing at least one video stream of a geographic location, determining a high likelihood of an emergency incident occurring at the geographic location and determining an incident type of the emergency incident based upon comparing patterns in the analyzed video stream against previously defined patterns associated with different incidents, automatically providing an alert in a user interface of the emergency response system, wherein said alert indicates the determined emergency incident, the determined incident type, and presents the video stream including a portion of video upon which the patterns were determined.

8. The method of claim 1, further comprising:

equipping responders with at least a video capture device and a video display device that are communicatively linked to at least one of the emergency response system and the video operations center (VOC), wherein video from said video capture device is routed to at least one of the emergency response system and the video operations center (VOC) where the routed video is programmatically analyzed for patterns detection purposes, and wherein video from at least one of the emergency response system and video operations center (VOC) is selectively presented upon the video display device.

9. The method of claim 1, wherein the video analysis engine is configured to recognize patterns in video streams, to compare these patterns against a data store of previously established patterns that are associated with defined incidents to determine occurrences of instances depicted in the video stream, and wherein the emergency response system is configured to dispatch emergency personnel to a site of an incident in response to an occurrence of an incident, and is configured to receive emergency calls alerting personnel of an emergency requiring a dispatch of emergency personnel.

10. The method of claim 1, further comprising:

granting control of at least one video source proximate to the geographic location to at least one of the emergency response system and the video operations center (VOC), wherein the control permits adjustments to be remotely made to the video source, which alter a video stream acquired from the video source, said altered video stream being routed to at least one of the emergency response system and the video operations center (VOC), whereby control comprises one of video source angle, video source focal point, video source contrast, and zoom.

11. The method of claim 1, further comprising

dynamically adjusting video coverage during the incident in real-time based upon results of pattern detection to ensure a maximization of video related to the incident is captured and available to at least one of the emergency response system and the video operations center (VOC), wherein the pattern detection from the video analysis in conjunction with an ability to control video captured by remotely located video sources proximate to the geographic location ensures that important entities, which are identified based upon the video analysis, and that are involved in the incident are tracked by video routed to at least one of the emergency response system and the video operations center (VOC).

12. A computer program product for emergency response comprising a computer readable storage medium that is a physical storage medium having computer usable program code embodied therewith, said computer usable program code being configured to be utilized by a processor, the computer program product comprising:

computer usable program code configured to integrate video analysis capabilities into at least one of an emergency response system and a video operations center (VOC) monitoring for emergency incidents; and

computer usable program code configured to perform at least one computer program driven response action involving detecting patterns in video streams that comprise video of an incident, wherein the response action comprises at least one of;

(A) responsive to a receipt of an emergency call by the emergency response system for an incident occurring at a geographic location, automatically executing a programmatic action to increase video capture activity at the geographic location and detecting patterns in video streams acquired from the increased video capture activity;

(B) analyzing at least one video stream of a geographic location and automatically dispatching an emergency responder to the geographic location based upon results of analyzing the at least one video stream, wherein the analyzing and automatic dispatching occur in at least one of real-time and near real-time; and

(C) analyzing at least one video stream of a geographic location, determining a high likelihood of an emergency incident occurring at the geographic location and determining an incident type of the emergency incident based upon comparing patterns in the analyzed video stream against previously defined patterns associated with different incidents, automatically providing an alert in a user interface of at least one of the emergency response system and the video operations center (VOC), wherein said alert indicates the determined emergency incident, the determined incident type, and presents the video stream including a portion of video upon which the patterns were determined.

13. The computer program product of claim 12, wherein said response action comprises:

responsive to a receipt of an emergency call by the emergency response system for an incident occurring at a geographic location, automatically executing a programmatic action to increase video capture activity at the geographic location and detecting patterns in video streams acquired from the increased video capture activity.

14. The computer program product of claim 12, wherein said response action comprises:

analyzing at least one video stream of a geographic location and automatically dispatching an emergency responder to the geographic location based upon results of analyzing the at least one video stream, wherein the analyzing and automatic dispatching occur in at least one of real-time and near real-time.

15. The computer program product of claim 12, further comprising:

analyzing at least one video stream of a geographic location, determining a high likelihood of an emergency incident occurring at the geographic location and determining an incident type of the emergency incident based upon comparing patterns in the analyzed video stream against previously defined patterns associated with different incidents, automatically providing an alert in a user interface of at least one of the emergency response system and the video operations center (VOC), wherein said alert indicates the determined emergency incident, the determined incident type, and presents the video stream including a portion of video upon which the patterns were determined.

16. The computer program product of claim 12, wherein the video analysis engine is configured to recognize patterns in video streams, to compare these patterns against a data store of previously established patterns that are associated with defined incidents to determine occurrences of instances depicted in the video stream, and wherein the emergency response system is configured to dispatch emergency personnel to a site of an incident in response to an occurrence of an incident, and is configured to receive emergency calls alerting personnel of an emergency requiring a dispatch of emergency personnel.

17. An emergency response system comprising:

a video analysis engine configured to recognize patterns in video streams, to compare these patterns against a data store of previously established patterns that are associated with defined incidents to determine occurrences of instances depicted in the video stream; and

an emergency response system configured to dispatch emergency personnel to a site of an incident in response to an occurrence of an incident, and is configured to receive emergency calls alerting personnel of an emergency requiring a dispatch of emergency personnel, wherein real-time results from the video analysis engine are integrated with the emergency response system, wherein results from the video analysis engine are used to automatically dispatch responders to an incident or are used to automatically alert a human agent of at least one of a video operations center and an emergency response center of a potential incident determined by the video analysis engine, wherein the human agent utilizes a computer linked to the emergency response system.

18. The system of claim 17, wherein the video analysis engine is further configured to selectively and automatically generate and convey an alert notification responsive to a detection of patterns indicative of an emergency incident, said system further comprising:

a pattern library able to store a plurality of information associated with a pattern, wherein said previously established patterns are digitally encoded within a physical storage medium of the pattern library; and

an emergency response interface configured to transmit and receive the alert notification, wherein the emergency response interface is a graphical user interface for the emergency response system usable by emergency response personnel.

19. The system of claim 18, further comprising:

a trigger analysis engine configured to automatically communicate the alert notification to a remotely located emergency response resource, wherein different emergency response resources are able to be registered with the trigger analysis engine to receive alter notifications satisfying a set of configurable conditions, wherein each emergency response resource is a computing device associated with a unique identifier.

20. The system of claim 18, wherein said video analysis engine is self training based upon feedback from recorded instances stored in a database of the emergency response system and based upon video associated with the recorded instances, wherein the self training permits the video analysis engine to create new patterns, which are stored in the pattern library and associated with incident specific elements.