METHODS, DEVICES AND SYSTEMS FOR DETECTING OBJECTS IN A VIDEO
First Claim
Patent Images
1. A method of detecting human objects in a video, comprising:
- determining pixels of a video image are foreground pixels, the group of foreground pixels constituting a foreground blob set of one or more foreground blobs;
for each of N locations within the video image, where N is an integer, comparing a corresponding predetermined shape with the foreground blob set to obtain a corresponding probability of a human at the location, thereby obtaining N probabilities corresponding to the N locations;
using the N probabilities, determining X humans are represented by the foreground blob set, where X is whole number, andproviding at least one of a report, an alarm, and an event detection using the determination of the representation of X humans,wherein a size of the corresponding predetermined shape for each of the N locations is determined in response to calibration of the video system.
3 Assignments
0 Petitions
Accused Products
Abstract
Methods, devices and systems for performing video content analysis to detect humans or other objects of interest a video image is disclosed. The detection of humans may be used to count a number of humans, to determine a location of each human and/or perform crowd analyses of monitored areas.
16 Citations
32 Claims
-
1. A method of detecting human objects in a video, comprising:
-
determining pixels of a video image are foreground pixels, the group of foreground pixels constituting a foreground blob set of one or more foreground blobs; for each of N locations within the video image, where N is an integer, comparing a corresponding predetermined shape with the foreground blob set to obtain a corresponding probability of a human at the location, thereby obtaining N probabilities corresponding to the N locations; using the N probabilities, determining X humans are represented by the foreground blob set, where X is whole number, and providing at least one of a report, an alarm, and an event detection using the determination of the representation of X humans, wherein a size of the corresponding predetermined shape for each of the N locations is determined in response to calibration of the video system. - View Dependent Claims (2, 3, 4, 5, 6, 7, 9, 10, 11, 12, 13, 14, 15, 16, 17, 19, 20, 21, 22, 23)
-
-
8. (canceled)
-
18. (canceled)
-
24. A method of detecting human objects in a video, comprising:
-
determining pixels of a video image of a real world scene are foreground pixels, the group of foreground pixels constituting a foreground blob set of one or more foreground blobs; for each of N locations within the video image, where N is an integer, comparing a corresponding predetermined shape with the foreground blob set to determine X humans are represented by the foreground blob set, where X is whole number and a location of each of the X humans is determined as a location within a horizontal plane of the real world; and providing at least one of a report, an alarm, and an event detection when the crowd density exceeds the threshold, using the determination of the representation of X humans, wherein a size of the corresponding predetermined shape for each of the N locations is determined in response to calibration of the video system, wherein the video system is used to obtain the video image. - View Dependent Claims (25, 26, 27, 28, 31, 32)
-
-
29-30. -30. (canceled)
Specification