System and method for counting people near external windowed doors

• US 8,224,026 B2
• Filed: 12/08/2006
• Issued: 07/17/2012
• Est. Priority Date: 12/08/2005
• Status: Expired due to Fees

• Current Assignee: UTC Fire & Security Americas Corporation, Inc. (Carrier Global Corporation)
• Original Assignee: Lenel Systems International Incorporated (Raytheon Technologies Corporation d/b/a RTX)
• Inventors: Golan, Oren, Pikaz, Arie, Raj, Jeffrey Sanders
• Primary Examiner(s): Fitzpatrick, Atiba O
• Application Number:
  US12/086,064
• Publication Number:
  US 20100092030A1
• Time in Patent Office: 2,048 Days
• Field of Search: 382100-107, 348113-117, 37524003-24029
• US Class Current: 382/103
• CPC Class Codes: G06V 10/255 Detecting or recognising potential candidate objects based on visual cues, e.g. shapes
• CPC Class Codes: G06V 40/10 Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands

Abstract

A system for counting objects, such as people, is provided having a camera (22) for capturing video images along an image plane (25) in which two-dimensional shapes or design (28) are spaced along the image plane (25), and a computer system (14, 22) for receiving the images and detecting objects associated with change occurring in the images, and counting one of the detected objects when of the detected objects approximates a shape (31) associated with the object being counted (e.g., ellipse shape, to count a person) that fully or substantially blocks a portion said two-dimensional shapes or design (28) in the image associated with the detected object, and the detected object meeting other criteria for the object, such as size or compactness of detected change within the object. The system is especially useful for counting people near external windowed doors (19) of a building entranceway doors by discriminating between spurious light crossing the image plane and people.

Claims

1. A system for counting objects crossing an area comprising:

a camera for capturing images along an image plane of a surface that is in the area, the surface includes a plurality of detectable features on the surface, wherein the objects to be counted cross over said surface; and

a processor for receiving said images and detecting objects associated with change occurring in said images, and counting one of said detected objects at least in accordance with said one of said detected objects having a shape that indicates the detected object is to be counted if the one of said detected objects at least substantially blocks a portion said plurality of detectable features in one of said images associated with the detected object.

2. The system according to claim 1 wherein said processor counts said one of said detected objects also in accordance with one or more characteristics of the detected object.

3. The system according to claim 2 wherein said one or more characteristics are one or more shape properties of the detected object.

4. The system according to claim 3 wherein said one or more shape properties represent at least one of the size of the detected object or compactness of change occurring within one or more parts of the detected object.

5. The system according to claim 1 wherein said processor is capable of detecting a direction of movement along the surface of at least said one of said detected objects.

6. The system according to claim 5 wherein said processor provides said counting of said one of said detected objects in accordance with the detected direction of said one of detected object.

7. The system according to claim 1 wherein said one of said detected objects counted is a person.

8. The system according to claim 7 wherein said person is represented by a geometric shape comprising an ellipse.

9. The system according to claim 1 wherein said camera comprises a video camera.

10. The system according to claim 1 comprising a user interface enabling a user to define a region along said image plane in which said plurality of detectable features extend both within said region and outside of said region to define an effective region within which said detected objects are locatable.

24. The system according to claim 1 wherein said surface is a floor adjacent an externally windowed door.

25. The system according to claim 1 wherein said surface comprises a floor of a room, wherein said floor comprises one or more of flat or raised elements.

26. The system according to claim 25 wherein said raised elements comprise a rug.

27. The system according to claim 1 wherein said surface comprises a mat located on a floor.

11. A method for counting objects comprising the steps of:

capturing images along an image plane of a surface that objects to be counted cross over, said surface including a plurality of detectable features on the surface;

detecting an object in one of said captured images associated with detected change along said image plane in said one of said captured images;

determining whether said detected object represents an object of interest when the detected object approximates a shape that indicates the object is to be counted if the detected object at least substantially blocks a portion of said plurality of detectable features in said one of said captured images; and

counting the detected object in accordance with one or more characteristics of the detected object responsive to determining that the detected object is an object of interest that at least substantially blocks a portion of said plurality of detectable features.

12. The method according to claim 11 wherein said one or more characteristics represent at least one of size or compactness of a change associated with the detected object.

13. The method according to claim 12 comprising

determining a direction of movement of the detected object crossing said image plane, and counting the detected object in accordance with the determined direction of movement.

14. The method according to claim 11 wherein said captured images comprise successive image frames of pixels, and said determining step comprises:

producing a differences image by subtracting one current image frame from a previous image frame;

thresholding said differences image to provide a binary image to identify pixels associated with change in said current image;

extracting one or more connecting components of said pixels associated with change indicated by said binary image; and

approximating an ellipse for each of said one or more connecting components in said binary image in which each approximated ellipse represents one or more detected objects.

15. The method according to claim 11 wherein said captured images comprise successive image frames of pixels, and said determining step comprises:

producing a background image of said image plane;

producing a differences image by subtracting one current image frame from said background image;

thresholding said differences image to provide a binary image to identify pixels associated with change in said current image;

extracting one or more connecting components of said pixels associated with change indicated by said binary image; and

approximating an ellipse for each of said one or more connecting components in said binary image in which each approximated ellipse represents one or more detected objects.

16. The method according to claim 15 wherein said counting step comprises:

counting a person as crossing said image plane when said detected object corresponds to the approximated ellipse and blocks in said current image frame a selected portion said plurality of detectable features.

17. The method according to claim 11 wherein said detecting, determining and counting steps are carried out with respect to a region of said images along said surface.

18. The method according to claim 17 wherein said captured images comprise successive image frames of pixels, and said determining step comprises:

generating a shape image which identifies locations of pixels associated with said plurality of detectable features in said region using images from said camera when no moving objects are present along said image plane; and

comparing within boundaries of an approximated shape of each of said detected objects the blocking of pixels of the shape image in said one of said captured images to determine whether at least a selected portion of the plurality of detectable features is blocked in said one of said captured images.

19. The method according to claim 11 wherein said surface is a floor adjacent an externally windowed door.

20. The method according to claim 11 wherein said detected object is counted if the detected object is a person.

21. The method according to claim 20 wherein said detected object is approximated by a geometric shape and wherein the geometric shape is an ellipse if the detected object is a person.

22. The method according to claim 11 wherein said detecting, determining, and counting steps are carried out by a computer system.

23. The method according to claim 11, comprising

defining a region along said image plane in which said plurality of detectable features extend both within said region and outside of said region to define an effective region of said images within which said detecting step is carried out.

28. The method according to claim 15 wherein said counting step comprises:

counting a person as crossing said image plane when each said ellipse blocks in said current image frame a selected portion of said plurality of detectable features.