Method of counting objects in a monitored environment and apparatus for the same

• US 7,346,187 B2
• Filed: 10/10/2003
• Issued: 03/18/2008
• Est. Priority Date: 10/10/2003
• Status: Active Grant

• Current Assignee: Johnson Controls Tyco IP Holdings LLP (Johnson Controls International plc)
• Original Assignee: Intellivid Corporation (Johnson Controls International plc)
• Inventors: Buehler, Christopher J.
• Primary Examiner(s): Wu; Jingge
• Assistant Examiner(s): Patel; Jayesh A
• Application Number:
  US10/683,483
• Publication Number:
  US 20050078852A1
• Time in Patent Office: 1,621 Days
• Field of Search: 382/291, 348/135, 348/169
• US Class Current: 382/103
• CPC Class Codes: G06T 7/0004 Industrial image inspection
• CPC Class Codes: G06V 20/53 Recognition of crowd images, e.g. recognition of crowd congestion

Abstract

Methods and system for determining, a number of objects, without tracking each of the objects, in first and second fields-of-view. First and second video frames are received from first and second image sources. The image sources have the first and second fields-of-view, and the fields-of-view are known to overlap at least in part. The number of objects is determined based on the first and second video frames and the known overlap.

Claims

1. A computerized method of video analysis comprising:

a) receiving a plurality of first video frames, the first video frames being generated over a period time by a first image source having a first field of view;

b) receiving a plurality of second video frames, the second video frames being generated over a period of time by a second image source having a second field of view, the second field of view having a known overlap with at least part of the first field-of-view; and

c) sub-dividing the first and second fields-of-view into two or more image regions;

d) classifying each of the image regions in the first field-of-view overlapping with image regions in the second field of view as an overlapping image region, classifying each of the image regions in the second field-of-view overlapping with image regions in the first field of view as an overlapping image region, and classifying the remainder of the image regions as non-overlapping image regions; and

(e) determining a number of objects, without tracking each of the objects, in the first and second fields-of-view wherein the number of objects equals the sum of the number of objects included in each non-overlapping image region and the maximum number of objects among each of the overlapping image regions.

2. The computerized method of video analysis of claim 1 further comprising maintaining a data structure identifying the image regions as either overlapping or non-overlapping.

3. The computerized method of claim 2 further comprising:

receiving updated overlap information; and

altering data in the data structure identifying image regions that correspond to overlapping or non-overlapping image regions based on the updated overlap information.

4. The computerized method of claim 1 further comprising averaging the number of objects included in the first and second fields-of-view over at least two instants in time.

5. A computerized system for video analysis comprising:

a) a receiving module configured to receive a plurality of first video frames, the first video frames being generated over a period time by a first image source having a first field-of-view and to receive a plurality of second video frames, the second video frames being generated over a period of time by a second image source having a second field-of-view, the second field of view having a known overlap with at least part of the first field-of-view; and

b) a processing module configured to (i) sub-divide the first and second fields-of-view into two or more image regions, (ii) classify each of the image regions in the first field-of-view overlapping with image regions in the second field of view as an overlapping image region, (iii) classify each of the image regions in the second field-of-view overlapping with image regions in the first field of view as an overlapping image region, (iv) classify the remainder of the image regions as non-overlapping image regions; and

(v) determine a number of objects, without tracking the objects, wherein the number of objects equals the sum of the number of objects included in each non-overlapping image region and the maximum number of objects among each of the overlapping image regions.

6. The computerized system for video analysis of claim 5 wherein, the processing module is further configured to maintain a data structure identifying the image regions classified as either overlapping or non-overlapping.

7. The computerized system for claim 6, wherein the processing module is further configured to receive updated overlap information, and to alter data in the data structure overlapping or non-overlapping image regions based on the updated overlap information.

8. The computerized system for claim 5 wherein the processing module is further configured to average the number of objects included in the first and second fields-of-view over at least two instants in time.