System and method for video episode viewing and mining

1. A computer-implemented method of automatically analyzing video data to identify movement patterns, the method comprising:

receiving, from one or more cameras, video data comprising a plurality of frames representing images of one or more objects within a physical area, wherein:

the physical area is associated with a grid, the grid comprising a plurality of grid cells; and

each camera of the one or more cameras is configured to capture video of one or more grid cells of the grid;

identifying a plurality of events within the video data, wherein an event represents a movement of an object of interest from a first grid cell of the grid associated with the physical area to a second grid cell of the grid;

generating a plurality of event data records reflecting the plurality of events; and

determining one or more frequent episodes from the plurality of event data records, wherein an episode comprises a series of events associated with a particular object of interest.

2. The method of claim 1 , wherein the video data is captured from a plurality of cameras having different optical perspectives with respect to the physical area.

3. The method of claim 1 , wherein identifying an event comprises:

identifying a first object of interest in a first frame;

identifying a first grid position of the first object of interest in the first frame;

identifying a second object of interest in a second frame, wherein the second frame is adjacent to the first frame;

identifying a second grid position of the second object of interest in the second frame;

determining that the second object of interest matches the first object of interest; and

determining that the second grid position differs from the first grid position.

4. The method of claim 3 , further comprising:

generating an event data record reflecting the first grid position and the second grid position of a monitored object of interest, wherein the monitored object of interest corresponds to the first object of interest and the second object of interest.

5. The method of claim 4 , further comprising:

generating an object identifier that uniquely identifies the monitored object of interest; and

storing the object identifier in the event data record.

6. The method of claim 3 , wherein determining that the second object of interest matches the first object of interest comprises:

generating an object signature, wherein the object signature reflects distinct graphical characteristics of the first object of interest; and

determining that the second object of interest matches the object signature.

7. The method of claim 6 , wherein determining one or more frequent episodes comprises:

identifying a series of events such that, for each event, the second object matches the object signature within a specified tolerance range.

8. The method of claim 1 , further comprising:

identifying the object of interest by determining that a detected object in the video data is not a marker object.

9. The method of claim 1 , further comprising:

identifying the object of interest by determining that a detected object in the video data matches characteristics of a particular category of objects.

10. The method of claim 9 , wherein determining one or more frequent episodes comprises:

identifying a series of events such that, for each event, the object of interest matches characteristics of the particular category within a specified tolerance range.

11. The method of claim 1 , wherein the one or more frequent episodes represent one or more paths frequently traveled by humans in the physical area.

12. The method of claim 1 , wherein:

the physical area is a traffic area; and

the one or more frequent episodes represent one or more common traffic patterns in the traffic area.

13. A system configured to automatically analyze video data to identify movement patterns, the system comprising:

a processing system comprising one or more processors; and

a memory system comprising one or more computer-readable media, wherein the computer-readable media contain instructions stored thereon that, when executed by the processing system, cause the system to perform operations comprising:

receiving, from one or more cameras, video data comprising a plurality of frames representing images of one or more objects within a physical area, wherein:

the physical area is associated with a grid, the grid comprising a plurality of grid cells; and

each camera of the one or more cameras is configured to capture video of one or more grid cells of the grid;

identifying a plurality of events within the video data, wherein an event represents a movement of an object of interest from a first grid cell of the grid associated with the physical area to a second grid cell of the grid;

generating a plurality of event data records reflecting the plurality of events; and

determining one or more frequent episodes from the plurality of event data records, wherein an episode comprises a series of events associated with a particular object of interest.

14. The system of claim 13 , wherein the video data is captured from a plurality of cameras having different optical perspectives with respect to the physical area.

15. The system of claim 13 , wherein identifying an event comprises:

identifying a first object of interest in a first frame;

identifying a first grid position of the first object of interest in the first frame;

identifying a second object of interest in a second frame, wherein the second frame is adjacent to the first frame;

identifying a second grid position of the second object of interest in the second frame;

determining that the second object of interest matches the first object of interest; and

determining that the second grid position differs from the first grid position.

16. The system of claim 15 , the operations further comprising:

generating an event data record reflecting the first grid position and the second grid position of a monitored object of interest, wherein the monitored object of interest corresponds to the first object of interest and the second object of interest.

17. The system of claim 16 , the operations further comprising:

generating an object identifier that uniquely identifies the monitored object of interest; and

storing the object identifier in the event data record.

18. The system of claim 15 , wherein determining that the second object of interest matches the first object of interest comprises:

generating an object signature, wherein the object signature reflects distinct graphical characteristics of the first object of interest; and

determining that the second object of interest matches the object signature.

19. The system of claim 18 , wherein determining one or more frequent episodes comprises:

identifying a series of events such that, for each event, the second object matches the object signature within a specified tolerance range.

20. The system of claim 13 , the operations further comprising:

identifying the object of interest by determining that a detected object in the video data is not a marker object.

21. The system of claim 13 , the operations further comprising:

identifying the object of interest by determining that a detected object in the video data matches characteristics of a particular category of objects.

22. The system of claim 21 , wherein determining one or more frequent episodes comprises:

identifying a series of events such that, for each event, the object of interest matches characteristics of the particular category within a specified tolerance range.

23. The system of claim 13 , wherein the one or more frequent episodes represent one or more paths frequently traveled by humans in the physical area.

24. The system of claim 13 , wherein:

the physical area is a traffic area; and

the one or more frequent episodes represent one or more common traffic patterns in the traffic area.

25. A computer-implemented method of automatically analyzing video data to identify movement patterns, the method comprising:

receiving, from one or more cameras, video data comprising a plurality of frames representing images of one or more objects within a physical area, wherein:

the physical area is associated with a grid, the grid comprising a plurality of grid cells; and

each camera of the one or more cameras is configured to capture video of one or more grid cells of the grid;

identifying a plurality of events, wherein identifying an event comprises:

identifying a first object of interest in a first frame;

identifying, within a first grid cell of the grid, a first grid position of the first object of interest in the first frame;

identifying a second object of interest in a second frame, wherein the second frame is adjacent to the first frame;

identifying, within a second grid cell of the grid, a second grid position of the second object of interest in the second frame;

determining that the second object of interest matches the first object of interest;

determining that the second grid position differs from the first grid position, based on identifying that the first grid position is within the first grid cell and that the second grid position is within the second grid cell; and

generating an event data record reflecting the first grid position and the second grid position of a monitored object of interest, wherein the monitored object of interest corresponds to the first object of interest and the second object of interest; and

determining one or more frequent episodes from the generated event data records, wherein an episode comprises a series of events associated with a monitored object of interest.