System and method for video content analysis using depth sensing
First Claim
Patent Images
1. A video content analysis method comprising:
- receiving a video sequence that includes a plurality of frames, each frame including a video image;
for each frame, receiving two-dimensional (2D) image data and also receiving depth data;
processing the 2D image data of the video sequence to differentiate foreground data from background data and to detect one or more blobs comprised of the foreground data, the one or more blobs corresponding to one or more objects, wherein differentiating the foreground data from the background data is performed without analyzing the depth data;
for each detected blob, using the depth data to determine whether at least part of the blob corresponds to at least part of a target by at least (1) mapping the blob to a set of Z-planes;
(2) determining that on at least some Z-planes the blob is clustered into different blob regions corresponding to two objects separated in space; and
(3) grouping the separated blob regions of the Z-planes into two physical objects by checking their spatial overlaps, wherein one of the physical objects corresponds to the target; and
after it is determined that at least part of a blob corresponds to at least part of a target, tracking the target and detecting at least one event associated with the target.
5 Assignments
0 Petitions
Accused Products
Abstract
A method and system for performing video content analysis based on two-dimensional image data and depth data are disclosed. Video content analysis may be performed on the two-dimensional image data, and then the depth data may be used along with the results of the video content analysis of the two-dimensional data for tracking and event detection.
-
Citations
22 Claims
-
1. A video content analysis method comprising:
-
receiving a video sequence that includes a plurality of frames, each frame including a video image; for each frame, receiving two-dimensional (2D) image data and also receiving depth data; processing the 2D image data of the video sequence to differentiate foreground data from background data and to detect one or more blobs comprised of the foreground data, the one or more blobs corresponding to one or more objects, wherein differentiating the foreground data from the background data is performed without analyzing the depth data; for each detected blob, using the depth data to determine whether at least part of the blob corresponds to at least part of a target by at least (1) mapping the blob to a set of Z-planes;
(2) determining that on at least some Z-planes the blob is clustered into different blob regions corresponding to two objects separated in space; and
(3) grouping the separated blob regions of the Z-planes into two physical objects by checking their spatial overlaps, wherein one of the physical objects corresponds to the target; andafter it is determined that at least part of a blob corresponds to at least part of a target, tracking the target and detecting at least one event associated with the target. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11)
-
-
12. A video content analysis method comprising:
-
receiving a video sequence that includes a plurality of frames, each frame including a video image; for each frame, receiving two-dimensional (2D) image data and also receiving depth data; processing the 2D image data of the video sequence to differentiate foreground data from background data and to detect one or more blobs comprised of the foreground data, the one or more blobs corresponding to one or more objects, wherein differentiating the foreground data from the background data is performed without analyzing the depth data; for each detected blob, using the depth data to determine whether to track at least a first part of the blob as a target; and after it is determined to track the target, detecting at least one event associated with the target, wherein determining whether to track at least the first part of the blob as a target includes; mapping the blob to a set of Z-planes; determining that on at least some Z-planes the blob is clustered into different blob regions corresponding to the first part of the blob and a second part of the blob separated in space; and grouping blob slices corresponding to the Z-planes from the first part of the blob to correspond to a physical object by checking their spatial overlaps, wherein the physical object corresponds to the target. - View Dependent Claims (13, 14, 15, 16, 17, 18, 19, 20, 21, 22)
-
Specification