Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
2002-06-18
2004-04-27
Lee, Richard (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
Reexamination Certificate
active
06728313
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a method and apparatus for operating a video decoder at an increased rate of speed. More particularly, the present invention relates to a method and apparatus for performing dequantization and Inverse Discrete Cosine Transform (IDCT) on video signal data in a video decoder at a rate of speed compatible with a 30 frames per second motion picture quality.
2. The Background Art
Graphics and video processing are operation intensive. At the same time, highspeed processing is particularly important in the areas of video processing, image compression and decompression. Furthermore, with the growth of the “multi-media” desktop, it is imperative that computer systems accommodate high-speed graphics, video processing, and image compression/decompression to execute multimedia applications. Accordingly, it would be desirable if a video decoder were designed to maintain a speed compatible with a 30 frames per second motion picture quality.
Video decoding includes the steps of dequantization, IDCT, motion compensation, and color space conversion. Each picture, or frame, processed by the video decoder comprises a plurality of macroblocks, each of which further comprise a plurality of blocks of encoded video signal data. Dequantization is performed on each block of encoded video signal data, and produces an 8×8 matrix corresponding to each block. Since IDCT typically includes multiplication of each of these 64 dequantization values by a cosine matrix, the IDCT process is particularly time-consuming, and a bottleneck of the speed of the decoder.
The speed of the decoder is limited by the speed of the IDCT process. Typically, as many as 10 multiplications are required to complete one IDCT row or column calculation. For a resolution of 640×480, the number of blocks in each frame to be processed for a 4:2:0 format is 7200. Thus, the total number of calculations required to process one frame is 10*(8+8)*7200=1,152,000. Clearly, the number of calculations performed during the IDCT process substantially limits the speed of the decoder.
According to current standards, it would be desirable to maintain the quality of the decoder at 30 frames per second as required for the motion picture quality. Therefore, it would be beneficial if the speed of the IDCT process could be increased, thereby speeding up the decoding process. A need exists in the prior art for a method for performing the IDCT calculations at an increased rate of speed through reducing the number of IDCT calculations required.
BRIEF DESCRIPTION OF THE INVENTION
The present invention provides an improved method and apparatus for performing dequantization and IDCT calculations in an MPEG-II decoder. A dequantization block is provided for performing dequantization calculations on a block of encoded video signal data using a modified standard quantization matrix. The modified standard quantization matrix is a product of a standard quantization matrix and a diagonal cosine matrix. An IDCT block is provided for performing IDCT calculations on each block processed by the dequantization block. Through combination of the standard quantization and diagonal cosine matrices prior to the IDCT process, the number of operations required during the IDCT process is substantially reduced.
The dequantization block receives a modified standard quantization matrix, the modified standard quantization matrix being a product of a standard quantization matrix corresponding to the encoded video data stream and a diagonal cosine matrix. In addition the dequantization block receives a scale representing a compression ratio of the encoded video data stream and a non-zero IDCT coefficient matrix corresponding to a block of the encoded video data. The dequantization block then multiplies the scale, the non-zero IDCT coefficient matrix and the modified standard quantization matrix to produce dequantization video signal data.
The IDCT block receives each block of processed data from the dequantization block. The IDCT block then performs IDCT row and column calculations on the dequantization video signal data according to a set of IDCT butterfly operations.
The present invention includes a dequantization block and an IDCT block which operate in parallel to maximize the speed of the MPEG-II decoder. Through the movement of multiplication of a cosine matrix typically performed during the IDCT process to the prior dequantization step, the remaining steps in the IDCT process are recombined to reduce the total number of operations required by the IDCT block. As a result, the total number of operations performed during decoding is substantially reduced, and the speed of the decoding process is correspondingly increased.
REFERENCES:
patent: 5065346 (1991-11-01), Kawai et al.
patent: 5136385 (1992-08-01), Campbell
patent: 5357282 (1994-10-01), Lee
patent: 5410352 (1995-04-01), Watanabe
patent: 5418962 (1995-05-01), Bodin et al.
patent: 5428456 (1995-06-01), Parulski et al.
patent: 5467131 (1995-11-01), Bhaskaran et al.
patent: 5477242 (1995-12-01), Thompson et al.
patent: 5523847 (1996-06-01), Feig et al.
patent: 5592194 (1997-01-01), Nishikawa
patent: 5699117 (1997-12-01), Uramoto et al.
patent: 5754235 (1998-05-01), Urano et al.
patent: 5768429 (1998-06-01), Jabbi et al.
patent: 5777679 (1998-07-01), Cheney et al.
patent: 5825680 (1998-10-01), Wheeler et al.
patent: 5844609 (1998-12-01), Filor et al.
patent: 5903312 (1999-05-01), Malladi et al.
patent: 5909254 (1999-06-01), Feig et al.
patent: 5936669 (1999-08-01), Niesen
Hsu Roger
Hu Xiaoping
Liu Yan
Ouyang Jack
Tu Kuei-Chung
Hanish Marc
Lee Richard
Thelen Reid & Priest LLP
LandOfFree
Method and apparatus for performing MPEG II dequantization... 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 apparatus for performing MPEG II dequantization..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for performing MPEG II dequantization... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3197406