Image analysis – Pattern recognition – Classification
Reexamination Certificate
1999-12-01
2004-06-22
Patel, Jayanti K. (Department: 2625)
Image analysis
Pattern recognition
Classification
C348S169000, C348S699000, C382S103000, C382S118000, C382S156000, C382S159000, C382S176000, C382S203000
Reexamination Certificate
active
06754389
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to the field of communications and information processing, and in particular to the field of video categorization and retrieval.
2. Description of Related Art
Consumers are being provided an ever increasing supply of information and entertainment options. Hundreds of television channels are available to consumers, via broadcast, cable, and satellite communications systems. Because of the increasing supply of viewing options, it is becoming increasingly more difficult for a consumer to locate programs of specific interest. A number of techniques have been proposed for easing the selection task, most of which are based on a classification of the available programs, based on the content of each program.
A classification of program material may be provided via a manually created television guide, or by other means, such as an auxiliary signal that is transmitted with the content material. Such classification systems, however, are typically limited to broadcast systems, and require the availability of the auxiliary information, such as the television guide or other signaling. Additionally, such classification systems do not include detailed information, such as the time or duration of commercial messages, news bulletins, and so on. A viewer, for example, may wish to “channel surf” during a commercial break in a program, and automatically return to the program when the program resumes. Such a capability can be provided with a multi-channel receiver, such as a picture-in-picture receiver, but requires an identification of the start and end of each commercial break. In like manner, a viewer may desire the television to remain blank and silent except when a news or weather bulletin occurs. Conventional classification systems do not provide Sufficient detail to support selective viewing of segments of programs.
Broadcast systems require a coincidence of the program broadcast time and the viewer's available viewing time. Video recorders, including multiple-channel video recorders, are often used to facilitate the viewing of programs at times other than their broadcast times. Video recorders also allow viewers to select. specific portions of recorded programs for viewing. For example, commercial segments may be skipped while viewing an entertainment or news program, or, all non-news material may be skipped to provide a consolidation of the day's news at select times. Conventional classification systems are often incompatible with a retrieval of the program from a recorded source. The conventional television guide, for example, provides information for locating a specific program at a specific time of day, but cannot directly provide information for locating a specific program on a recorded disk or tape. As noted above, the conventional guides and classification systems are also unable to locate select segments of programs for viewing.
BRIEF SUMMARY OF THE INVENTION
It is an object of this invention to provide a method and system that facilitate an automated classification of content material within segments, or clips, of a video broadcast or recording. The classification of each segment within a broadcast facilitates selective viewing, or non-viewing, of particular types of content material, and can also be used to facilitate the classification of a program based on the classification of multiple segments within the program.
The object of this invention, and others, are achieved by providing a content-based classification system that detects the presence of objects within a frame and determines the path, or trajectory, of each object through multiple frames of a video segment. In a preferred embodiment, the system detects the presence of facial images and text images within a frame and determines the path, or trajectory, of each image through multiple frames of the video segment. The combination of face trajectory and text trajectory information is used in a preferred embodiment of this invention to classify each segment of a video sequence. To enhance the classification process, a hierarchical information structure is utilized. At the upper, video, information layer, the parameters used for the classification process include, for example, the number of object trajectories of each object type within the segment, an average duration for object type trajectory, and so on. At the lowest, model, information layer, the parameters include, for example, the type, color, and size of the object corresponding to each object trajectory. In an alternative embodiment, a Hidden Markov Model (HMM) technique is used to classify each segment into one of a predefined set of classifications, based on the observed characterization of the object trajectories contained with the segment.
REFERENCES:
patent: 5710875 (1998-01-01), Harashima et al.
patent: 5812193 (1998-09-01), Tomitaka et al.
patent: 5912980 (1999-06-01), Hunke
patent: 5983251 (1999-11-01), Martens et al.
patent: 6028956 (2000-02-01), Shustorovich et al.
patent: 6072542 (2000-06-01), Wilcox et al.
patent: 6118887 (2000-09-01), Cosatto et al.
patent: 6275614 (2001-08-01), Krishnamurthy et al.
patent: 6404925 (2002-06-01), Foote et al.
patent: 6445810 (2002-09-01), Darrell et al.
patent: 6456328 (2002-09-01), Okada
patent: 0005918 (1979-12-01), None
patent: 0805405 (1997-05-01), None
patent: 0805405 (1997-11-01), None
patent: WO9427404 (1994-11-01), None
patent: WO9740454 (1997-10-01), None
patent: WO9748068 (1997-12-01), None
Zhong et al: “Video model and segmentation for content-based video indexing,” Circuits And Systems, 1997, ISCAS ′97., Proceedings of 1997 IEEE International Syumposium On Hong Kong Jun. 9-12, 1997, Jun. 9, 1997, pp. 1492-1495.
Pavlovic et al: “Time-Series Classification Using Mixed-State Dynamic Bayesian Networks,” Proceedings 1999 IEEE Computer Society Conference On Computer Vision And Pattern Recognition. Jun. 23-25, 1999, Jun. 23, 1999, pp. 609-615.
Agnihotri Lalitha
Dimitrova Nevenka
Wei Gang
Desire Gregory
Patel Jayanti K.
LandOfFree
Program classification using object tracking does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Program classification using object tracking, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Program classification using object tracking will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3361282