Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
1999-04-02
2001-05-15
Britton, Howard (Department: 2713)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
C375S240120, C382S236000
Reexamination Certificate
active
06233277
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to decoding of digitally encoded video signals and, more particulary, to a video decoder and method for decoding compressed high-definition video data.
BACKGROUND OF THE INVENTION
Digital encoding and decoding of high-definition video signals provides higher image resolution and more effective control of the storage, manipulation and display of the video signal over existing analog NTSC and PAL video systems. In an HDTV environment, for example, digital video data is encoded at a transmission end in accordance with a specified compressed bitstream syntax, such as the MPEG-2 standard, and decoded at a receiving end in accordance with a specified decoding process. The decoded video signal is reconstructed at the receiving end into picture frames that may be presented for display with an HDTV resolution of 1920×1080 or reduced to a standard NTSC resolution of 720×480.
Decoding of compressed video signals is a memory-intensive process, especially for compressed high-definition video signals. Receivers adapted to decode compressed high-definition video signals generally require a significant amount of memory to store reference frame data and additional side information required in the decoding process. In the past, for example, a typical HDTV receiver for decoding a high-definition MPEG-2 video bitstream has required 12 MB or more of random-access-memory (RAM) to provide adequate memory storage of control information and reference frame video data in the receiver for reconstructing the transmitted picture frames.
Due to the relatively high cost of high speed memory, developers of HDTV receivers have sought to reduce the amount of memory required in the receiver for the decoding process. In the past, one approach has been to exploit only a portion of the compressed data in the transmitted HDTV signal to produce a decoded video signal representing an image of lesser resolution. For example, HDTV receivers have been designed in the past that decimate the decoded video data in accordance with a predetermined decimation scheme either before or after the inverse discrete transform function of the decoder to reduce the amount of video data that is stored in memory for reconstruction of the picture frames.
Decoded DCT coefficients may be decimated by masking a block of DCT coefficients of an 8×8 DCT coefficient array before the remaining coefficients are applied to the IDCT circuit of the decoder. Alternatively, particular rows and columns of pixel data generated by the IDCT circuit in an 8×8 pixel data array may be eliminated to reduce the amount of video data that must be stored. The reconstructed picture frames are then displayed at a lower NTSC resolution.
While this approach requires less memory in the decoder for reconstructing picture frames, the decimation scheme performed by the decoder may result in reduced picture quality as a portion of the decoded video data is eliminated during the decoding process. Furthermore, since only a limited portion of the decoded video data can be eliminated without completely sacrificing picture quality, the decimation scheme achieves only a modest amount of compression and is not readily adaptable for memory-scalable applications where the amount of decoder memory may vary among different decoding applications.
One technique for conserving memory is to use DCT based compression and decompression as described in U.S. patent application Ser. No. 09/178,980, entitled “Reduced Memory Video Decoder for Compressed High-Definition Video Data”, filed on Oct. 26, 1998, and assigned to the assignee of this application. DCT compression is a very accurate compression process that has minimal data loss; however, DCT compression is relatively complicated. If the computational resources are capable of processing a DCT compression, then DCT compression should be used. However, if the computational resources are limited and not sufficient to process DCT compression, then other processes must be found to down-convert the data for storage.
Thus, there is a need for a video decoder and method for decoding compressed high-definition video data that provides visibly good quality reconstructed picture frames for relatively small memory requirements, for example, 2 megabytes or less.
SUMMARY OF THE INVENTION
The present invention overcomes the foregoing and other shortcomings and drawbacks of compressed video signal decoders and decoding methods heretofore known. The video decoder of the present invention permits substantial down-converting of video data with limited processing power into a memory. Further, in one embodiment, the video decoder of the present invention is sensitive to different picture complexities and continuously modifies the down-converting and up-converting process to optimize picture quality. Thus, the video decoder of the present invention provides better picture quality than other known decoding processes where the processor is limited.
In accordance with the principles of the present invention, a video decoder includes a macroblock (MB) parser and VLD circuit for parsing the incoming compressed video bitstream and decoding the block-level quantized DCT coefficients and motion vectors that are required in the frame or field reconstruction process. The decoded DCT coefficients are applied to an inverse quantization (IQ) and inverse Discrete Cosine Transform circuit to generate pel values (I-frames) or pel prediction errors (P- and B-frames) for each block of video data. The decoded motion vectors are applied to a motion compensator circuit. Reconstructed I- and P-reference frame video data are reduced in a down-converter by decimating and low-pass filtering to provide down-converted data for storage in the memory. During the decoding of predictive P- and B-frames, corresponding reference frame video data is retrieved from frame buffer memory and applied to an up-converter which interpolates and low-pass with high boost filters the down-converted data to generate corresponding up-converted data. The up-converter provides data from I- and P-reference frames on-the-fly for use in motion compensated prediction.
In accordance with another embodiment of the invention, the down-converter and the up-converter are adaptive in that the slope of the filter cutoff between sharp and gradual cutoff values is chosen as a function of a video signal-noise complexity.
In a further embodiment of the invention, a method of down-converting by decimating and adaptive low-pass filtering and up-converting by interpolating and adaptive low-pass with high boost filtering is provided utilizing a spatio-temporal adaptive Gaussian filter.
The above and other objects and advantages of the present invention shall be made apparent from the accompanying drawings and the description thereof.
REFERENCES:
patent: 5451949 (1995-09-01), Gundry
patent: 5627601 (1997-05-01), Ran et al.
patent: 5777677 (1998-07-01), Linzer et al.
patent: 5777812 (1998-07-01), Kim
patent: 5818530 (1998-10-01), Canfield et al.
patent: 5844614 (1998-12-01), Chong et al.
patent: 5862266 (1999-01-01), Hunter
patent: 6005983 (1999-12-01), Anderson et al.
Decoder Complexity And Performance Comparison Of Matching Pursuit And DCT-Based Video Codecs, Ralph Neff, Toshio Nomura, Avideh Zakhor, IEEE, pp. 783-787, Jan. 1998.
Hoang Dzung Tien
Ozcelik Taner
Britton Howard
Diep Nhon T
Sony Corporation
Wood Herron & Evans L.L.P.
LandOfFree
Reduced-memory video decoder for compressed high-definition... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Reduced-memory video decoder for compressed high-definition..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Reduced-memory video decoder for compressed high-definition... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2472020