Intelligent video analysis system and method

1. A method of analyzing a video, performed by a video analysis system, the method comprising:

in response to a camera detecting an event, selecting at least one logical camera group from a plurality of cameras, the selecting being performed by a logical camera group generating processor;

setting ranks for cameras of the at least one logical camera group, respectively, based on event metadata related to the event, the setting being performed by an intelligent rank setting processor; and

performing load balancing with respect to data transmitted from the cameras of the at least one logical camera group based on the set ranks, the load balancing being performed by a video analysis server that performs video analysis,

wherein the setting of the ranks comprises setting different ranks for cameras that are in a same logical camera group,

wherein the performing of the load balancing comprises assigning more resources of the video analysis system to data processing of a video received from a first camera than data processing of a video received from a second camera having a lower rank than the first camera, and

wherein the selecting of the at least one logical camera group further comprises:

grouping cameras that are physically connected to the camera that detected the event as a physically connected camera group; and

logically selecting cameras, which are related to the event, from among the physically connected camera group, based on the event metadata.

2. A method of analyzing a video, performed by a video analysis system, the method comprising:

in response to a camera detecting an event, selecting at least one logical camera group from a plurality of cameras, the selecting being performed by a logical camera group generating processor;

setting ranks for cameras of the at least one logical camera group, respectively, based on event metadata related to the event, the setting being performed by an intelligent rank setting processor; and

performing load balancing with respect to data transmitted from the cameras of the at least one logical camera group based on the ranks, the load balancing being performed by a video analysis server that performs video analysis,

wherein the setting of the ranks comprises setting different ranks for cameras that are in a same logical camera group,

wherein the performing of the load balancing comprises assigning more resources of the video analysis system to data processing of a video received from a first camera than data processing of a video received from a second camera having a lower rank than the first camera, and

wherein the camera analyzes at least one captured video for an event detection, and the video analysis server performs further analysis based on the at least one captured video based on the event metadata.

3. The method of claim 1 , wherein when the event is related to a moving object, the event metadata comprises information about at least one of a type of the event, a time when the event occurs, a location where the event occurs, a direction and speed of the moving object, and an attribute of the moving object.

4. The method of claim 1 , further comprising assigning priorities to the at least one logical camera group, respectively, based on the event metadata, and

wherein a camera included in a logical camera group having a higher priority is assigned more resources for processing data than a camera included in a logical camera group having a lower priority.

5. The method of claim 1 , wherein the performing the load balancing comprises not receiving or not processing data transmitted from a camera to which a rank below a predetermined rank among the ranks is set, by the video analysis system.

6. The method of claim 2 , wherein the intelligent rank setting processor selects an algorithm for the further analysis, by the video analysis server, of the at least one captured video based on the event metadata.

7. The method of claim 1 , wherein elements of the event metadata are given different weights, and wherein the load balancing is performed further based on the different weights.

8. The method of claim 2 , wherein the cameras of the at least one logical camera group, which are located within a predetermined radius from the camera that has detected the event, are further grouped inside a same logical group.

9. A video analysis system for analyzing a video, the video analysis system comprising:

a logical camera group generating processor configured to, in response to a camera detecting an event, selecting at least one logical camera group from a plurality of cameras;

an intelligent rank setting processor configured to set ranks to cameras of the at least one logical camera group, respectively, based on event metadata related to the event; and

a load balancing processor configured to perform load balancing with respect to data transmitted from the cameras of the at least one logical camera group based on the set ranks,

wherein the load balancing processor is further configured to assign more resources of the video analysis system to data processing of a video received from a first camera than data processing of a video received from a second camera having a lower rank than the first camera, and

wherein the logical camera group generating processor is further configured to group cameras that are located within a predetermined radius from the camera that detected the event as a physically connected camera group, and logically select cameras, which are related to the event, from among the physically connected camera group, based on the event metadata.

10. The system of claim 9 , wherein when the event is related to a moving object, the event metadata comprises information about at least one of a type of the event, a time when the event occurs, a location where the event occurs, a direction and speed of the moving object, and an attribute of the moving object.

11. The system of claim 9 , wherein the logical camera group generating processor is further configured to assign priorities to the at least one logical camera group, respectively, based on the event metadata, and

wherein the load balancing processor is further configured to assign more resources for processing data to a camera included in a logical camera group having a higher priority than a camera included in a logical camera group having a lower priority.

12. The system of claim 9 , wherein the system is configured to not receive or not process data transmitted from a camera to which a rank below a predetermined rank among the ranks is set for analyzing the video.

13. The system of claim 9 , wherein when the event is related to a moving object, the event metadata comprises information about at least one of a type of the event, a time when the event occurs, a location where the event occurs, a direction and speed of the moving object, and an attribute of the moving object.

14. The system of claim 9 , wherein elements of the event metadata are given different weights, and

wherein the load balancing processor is configured to perform the load balancing further based on the different weights.

15. A video surveillance method comprising:

receiving information including event metadata about an event of a moving object that is detected by at least one camera among a plurality of cameras, the event metadata indicating event type, event time, event location, and a direction and speed of the moving object;

generating at least one logical camera group by selecting cameras having a physical network connection with the at least one camera, selecting cameras located within a predetermined radius from the at least one camera, or selecting cameras located within a same surveillance zone as the at least one camera;

assigning ranks to cameras of the at least one logical camera group based upon respective priority values of the cameras of the at least one logical camera group with respect to elements of the event metadata and corresponding weight values of the elements of the event metadata; and

dynamically allocating resources to the cameras of the at least one logical camera group to achieve network load balancing by changing a number of video slots to be assigned to the plurality of cameras based upon the assigned ranks.

16. The method of claim 15 , wherein the dynamically allocating comprises allocating to a camera of a higher rank more resources for processing data than a camera of a lower rank.

17. The method of claim 15 , wherein the assigning comprises setting different ranks to cameras in a same logical camera group based on a direction of movement of the moving object.

18. The method of claim 15 , wherein the dynamically allocating comprises not receiving or not processing data transmitted from a camera to which a rank below a predetermined rank among the ranks is set.

19. A video surveillance method comprising:

detecting an event of a moving object by at least one camera among a plurality of cameras;

sending information including event metadata about the event, the event metadata indicating event type, event time, event location, and a direction and speed of the moving object; and

receiving a resource allocation,

wherein the allocation is performed by dynamically allocating resources based upon ranks assigned to cameras of at least one logical camera group,

wherein the at least one logical camera group is selected from the plurality of cameras by selecting cameras having a physical network connection with at least one camera, cameras being within a certain radius from the at least one camera, or cameras being within a same surveillance zone as the at least one camera, and

wherein the ranks are respectively assigned to the cameras of the at least one logical camera group based on respective priority values of the cameras of the at least one logical camera group with respect to elements of the event metadata and corresponding weight values of the elements of the event metadata.

20. The method of claim 19 , wherein more resources for processing data is allocated to a camera of a higher rank than a camera of a lower rank.

21. The method of claim 19 , wherein the elements of the event metadata are given different weight values, and wherein the resource allocation is performed further based on the different weight values.