Image analysis – Image compression or coding – Interframe coding
Reexamination Certificate
1998-02-18
2003-02-25
Johnson, Timothy M. (Department: 2623)
Image analysis
Image compression or coding
Interframe coding
C382S243000
Reexamination Certificate
active
06526173
ABSTRACT:
TECHNICAL FIELD
The present invention relates to methods and systems for encoding and decoding digital video data.
BACKGROUND ART
Video compression systems are employed to reduce the number of bits needed to transmit and store a digital video signal. As a result, a lower bandwidth communication channel can be employed to transmit a compressed video signal in comparison to an uncompressed video signal. Similarly, a reduced capacity of a storage device, which can comprise a memory or a magnetic storage medium, is required for storing the compressed video signal. A general video compression system includes an encoder, which converts the video signal into a compressed signal, and a decoder, which reconstructs the video signal based upon the compressed signal.
In the design of a video compression system, an objective is to reduce the number of bits needed to represent the video signal while preserving its visual content. Current methods and systems for video compression have achieved a reasonable quality of content preservation at a transmission bit rate of 56 kilobits per second. These methods and systems are based upon directly compressing a waveform representation of the video signal.
Motion compensation is one approach which is utilized in many video compression schemes. Current approaches model motion in terms of simple displacements of blocks or a global transformation of an entire scene to model camera motion.
SUMMARY OF THE INVENTION
The need exists for a video compression system which significantly reduces the number of bits needed to transmit and store a video signal, and which simultaneously preserves the visual content of the video signal.
It is thus an object of the present invention to significantly reduce the bit rate needed to transmit a video signal.
Another object of the present invention is to provide an efficient encoding of redundant temporal data contained within a digital video signal.
In carrying out the above objects, the present invention provides a method of encoding a video signal representative of a plurality of image frames. The plurality of image frames includes a starting frame, an ending frame, and at least one intermediate frame. In practice, high compression would be achieved by employing many frames of data between the starting frame and the ending frame. Included in the method is a step of identifying at least one object in the image frames, which can be performed such as by blue screen and chroma keying effects, or image segmentation either by hand or using automated signal processing techniques. During segmentation of the object, key points are identified. These key points are those that identify features on the object that should remain in essentially the same position relative to each other, for example, eyes, mouth, nose, ears, and hair line for a facial object.
For each object, a step of determining a mathematical transformation is performed. The mathematical transformation is determined by finding a single function that is applied to all the points in the object such that the function maps the selected key points from the starting frame to the ending frame. Since all intermediate frames can be constructed based on function, very high compression is achieved by the present invention because the encoded function requires an almost insignificant number of bits compared to the raw object data for the intermediate frames. As a consequence, the present invention provides a dynamic compression arrangement which produces correspondingly higher compression rates as the number of intermediate frames in the raw object data increases.
Further in carrying out the above objects, the present invention provides a method for decoding an encoded representation of a plurality of image frames. The plurality of image frames include a starting frame and at least one intermediate frame. The method includes a step of receiving an encoded signal containing an encoded representation for each of at least one object contained in a starting frame and of the same object contained in the ending frame of the image frames, and an encoded mathematical transformation for each of the at least one object. Each encoded representation is decoded to extract a representation for each of the at least one object in the starting frame and ending frame. Each encoded mathematical transformation is also decoded to extract a mathematical transformation for each of the at least one object. The representation for each object contained in the starting frame is transformed in accordance with the mathematical transformation corresponding thereto to form a transformed representation for each object in each of at least one intermediate frame. The transformation is used to map all of the points in the object in the starting frame to the intermediate frame and also to map all of the points in the ending frame to the intermediate frame. The points on the object in the intermediate frame are formed by combining the intermediate frame data derived from the starting frame with the intermediate frame data derived from the ending frame. The combination may be a linear or complex combination based upon the relative distance of the intermediate frame from the starting and ending frames. The transformed representation for each object is combined to form an intermediate frame. The steps of transforming and combining are repeated to form each intermediate frame.
Further in carrying out the above objects, systems are provided which perform the steps of the above-described methods.
Embodiments of the present invention advantageously produce high compression rates for coding object motion in video and film digital data. A practical temporal description of non-rigid, as well as rigid body motion for generic objects is provided based on actual object deformation rather than artificial motion descriptions based on arbitrary block matching techniques or even pixel matching techniques. This approach complements methods that provide a global description of camera motions such as zoom and pan.
These and other features, aspects, and embodiments of the present invention will become better understood with regard to the following description, appended claims, and accompanying drawings.
REFERENCES:
patent: 5253065 (1993-10-01), Richards et al.
patent: 5305400 (1994-04-01), Butera
patent: 5412435 (1995-05-01), Nakajima
patent: 5485611 (1996-01-01), Astle
patent: 5557684 (1996-09-01), Wang et al.
patent: 5623587 (1997-04-01), Bulman
patent: 5646675 (1997-07-01), Copriviza et al.
patent: 5703966 (1997-12-01), Astle
patent: 5831688 (1998-11-01), Yamada et al.
patent: 2 289 819 (1995-11-01), None
Publication entitledVideo Compression via Morphingby Lai Khen Siew, Nov. 21, 1995, pp. 1-19.
Beier, et al. “The Feature-Based Image Metamorphosis”, ACM-0-89791-479, 1992.
Crook John A.
Hughes Electronics Corporation
Johnson Timothy M.
Sales Michael W.
LandOfFree
Method and system for compression encoding video signals... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and system for compression encoding video signals..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and system for compression encoding video signals... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3129284