Interactive video system with dynamic video object descriptors
First Claim
1. A generator for generating a frame-sequential file of video object descriptors from successive frames of video information, comprising:
- object designation means for accepting an initial designation of video objects in a first video frame;
a video decomposition and object identification unit for decomposing subsequent frames of video information into constituent objects and for identifying objects therein;
an object tracker for tracking sequential positions of designated objects in sequential frames of video information;
an attribute generator for generating attributes corresponding to video objects tracked by said object tracker; and
a frame correlator for correlating frames of video information with objects identified by said video decomposition and identification unit;
whereby a frame-sequential file of video object descriptors is output, each such descriptor including, for each successive frame, a frame identifier correlated by said frame correlator and a video object descriptor for each object tracked by said object tracker.
2 Assignments
0 Petitions
Accused Products
Abstract
An interactive video system by which an operator is able to select an object moving in a video sequence and by which the interactive video system is notified which object was selected so as to take appropriate action. Interactive video is achieved through generation and use of video object descriptors which are synchronized to objects in the video sequence. Video object descriptors are generated by a generating tool which decomposes frames of video sequences and tracks movement of objects in those frames so as to generate a frame sequential file of video object descriptors. The file of video object descriptors are then used by an event interpreter which detects a match between the position of a pointing device on a display containing the video sequence and the position of a video object descriptor. When a match is detected, an interactive video operation is performed, such as jumping to a new video sequence, altering flow of the interactive video program or the like.
262 Citations
11 Claims
-
1. A generator for generating a frame-sequential file of video object descriptors from successive frames of video information, comprising:
-
object designation means for accepting an initial designation of video objects in a first video frame; a video decomposition and object identification unit for decomposing subsequent frames of video information into constituent objects and for identifying objects therein; an object tracker for tracking sequential positions of designated objects in sequential frames of video information; an attribute generator for generating attributes corresponding to video objects tracked by said object tracker; and a frame correlator for correlating frames of video information with objects identified by said video decomposition and identification unit; whereby a frame-sequential file of video object descriptors is output, each such descriptor including, for each successive frame, a frame identifier correlated by said frame correlator and a video object descriptor for each object tracked by said object tracker. - View Dependent Claims (2, 3, 4, 5, 6, 7)
-
-
8. A video event interpreter, comprising:
-
a video source for providing successive frames of video information; an object descriptor source for providing object descriptors that move in synchronism with movement of objects in the successive frames of video information; a display for displaying the successive frames of video information and on which a user-manipulable cursor is superimposingly displayed; an active object list for storing spatial attribute information for valid video objects, said active object list being extracted from said object descriptors; a cursor position matcher for matching the position of the user-manipulable cursor to spatial attribute information in the active object list; an action map containing a list of actions initiated in a case where said cursor position matcher matches the position of the cursor to a valid object in a list of active video objects; and means responsive to user selection of a video object within the video sequence for activating a corresponding action in said action map. - View Dependent Claims (9, 10, 11)
-
Specification