Hypervideo system and method with object tracking in a compressed digital video environment
First Claim
Patent Images
1. A method for object tracking in a digital video environment, comprising the steps of:
- identifying an object in a frame of the video; and
tracking an object, utilizing motion compensation information stored in the video, through at least one scene of the video to produce a hot region.
5 Assignments
0 Petitions
Accused Products
Abstract
A system and method for hypervideo object tracking in a compressed digital video environment includes a hypervideo authoring system and a hypervideo browser. The hypervideo authoring system permits one or more moving objects in an MPEG compressed video clip to be identified, tracked, refined, edited, and associated with a specified action with a minimum of user intervention. The hypervideo browser allows the MPEG video to be viewed and objects of interest to be selected by a user, thereby hyperlinking to additional information of interest.
-
Citations
28 Claims
-
1. A method for object tracking in a digital video environment, comprising the steps of:
-
identifying an object in a frame of the video; and
tracking an object, utilizing motion compensation information stored in the video, through at least one scene of the video to produce a hot region. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 28)
locating a point within the object; and
determining the boundaries of the object.
-
-
4. The method of claim 3, wherein the determining step comprises the substeps of:
-
approximating the contents of the frame; and
attributing all pixels contiguous to the point and within a range of color and brightness values to the object.
-
-
5. The method of claim 4, wherein the approximating step utilizes DC coefficients from color and brightness blocks in the frame.
-
6. The method of claim 1, wherein the tracking step comprises the substeps of:
-
identifying a current scene in which to track the object;
specifying the hot region in a first frame of the current scene to correspond to the object;
finding the object in each subsequent frame in the current scene.
-
-
7. The method of claim 6, wherein the finding step comprises the substeps of:
-
identifying a forward motion vector for each macroblock in the frame, wherein the forward motion vector corresponds to a forward reference area in a forward reference frame; and
copying any portion of the hot region from the forward reference area to the macroblock.
-
-
8. The method of claim 7, wherein the finding step further comprises the substeps of:
-
identifying a backward motion vector for each macroblock in the frame, wherein the backward motion vector corresponds to a backward reference area in a backward reference frame;
copying any portion of the hot region from the backward reference area to the macroblock; and
copying any portion of the hot region from the macroblock to the backward reference area.
-
-
9. The method of claim 6, wherein the tracking step further comprises filling holes in the hot region.
-
10. The method of claim 6, wherein the tracking step further comprises the substeps of:
-
locating scene transitions in the video;
determining whether a later scene corresponds to the end of the current scene; and
if so, continuing to track the object in the later scene.
-
-
11. The method of claim 10, further comprising the step of refining the hot region.
-
12. The method of claim 10, wherein the refining step comprises the substeps of:
-
identifying at least two contiguous hot subregions corresponding to the tracked object;
determining whether each of the hot subregions corresponds to the object;
if not, discarding a hot subregion; and
if so, merging the hot subregions into a merged hot region.
-
-
13. The method of claim 1, wherein the hot region is represented as a bitmap.
-
14. The method of claim 13, further comprising the step of converting the hot region to a polygon.
-
15. The method of claim 14, further comprising the step of storing the polygon in the video.
-
16. The method of claim 1, further comprising the step of specifying an action for the object.
-
17. The method of claim 16, further comprising the steps of:
-
playing the video;
waiting for a user to select an object; and
upon selection, performing the specified action.
-
-
18. The method of claim 16, wherein the video comprises a plurality of scenes and a default path through the scenes.
-
19. The method of claim 18, further comprising the step of identifying at least one alternate path through the scenes.
-
20. The method of claim 19, wherein the specified action comprises switching from the default path to the alternate path.
-
21. The method of claim 17, wherein the specified action comprises linking to additional information to be displayed.
-
22. The method of claim 1, further comprising the step of converting the hot region to a regular shape.
-
23. The method of claim 22, wherein the converting step preserves a specified shape.
-
24. The method of claim 23, wherein the converting step comprises the steps of:
-
locating a centroid for the hot region;
determining a size of the hot region; and
using the centroid and the size to place the regular shape.
-
-
28. The method of claim 1, wherein the tracking step comprises the substeps of:
-
locating scene transitions in the video;
determining whether a later scene corresponds to the end of the current scene; and
if so, continuing to track the object in the later scene.
-
-
25. A method for object tracking and editing in a digital video environment, comprising the steps of:
-
tracking an object through at least one scene of the video to produce a hot region;
identifying an erasure region of an object in a frame of the video;
editing the object by identifying and tracking the erasure region; and
erasing any portion of the hot region corresponding to the erasure region. - View Dependent Claims (26, 27)
locating a point within the object; and
determining the boundaries of the object.
-
Specification