Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
1999-07-21
2003-03-25
Kelley, Chris (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
Reexamination Certificate
active
06539056
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a picture decoding method and apparatus for decoding compressed picture data of a first resolution obtained on predictive coding by motion prediction in terms of a pre-set pixel block (macro-block) as a unit and on performing orthogonally-transform in terms of a pre-set pixel block (orthogonal transform block) as a unit. More particularly, it relates to a picture decoding method and apparatus for decoding compressed picture data of the first resolution and for decimating the data to moving picture data of a second resolution lower than the first resolution.
2. Description of the Related Art
There is now going on the standardization of digital television signals employing the picture compression system, such as Moving Picture Experts Group Phase 2 (MPEG2). Among the standards for digital television broadcast, there are a standard for standard resolution pictures, such as those with the number of effective lines in the vertical direction of 576, and a standard for high-resolution pictures, such as those with the number of effective lines in the vertical direction of 1152. Recently, there is raised a demand for a downdecoder for decoding compressed picture data of a high-resolution picture and for reducing the resolution of the compressed picture data by ½ to generate picture data of the picture data of standard resolution to display the picture data on a television monitor adapted to cope with the standard resolution.
There is proposed in a publication entitled “Scalable Decoder free of low-range Drift” (written by Iwahashi, Kanbayashi and Takaya, Shingaku-Gihou CS94-186, DSP 94-108, 1995-01) a downdecoder for decoding a bitstream of, for example, MPEG2, obtained on predictive coding with motion prediction of a high-resolution picture and compression coding by discrete cosine transform, and for downsampling the picture to a picture of standard resolution. This Publication, referred to below as Publication 1, shows the following first to third downdecoders.
Referring to
FIG. 1
, this first downdecoder includes an inverse discrete cosine transform unit
1001
, for processing a bitstream of a high resolution picture with 8 (number of coefficients as counted from the dc component in the horizontal direction)×8 (number of coefficients as counted from the dc component in the vertical direction), an adder
1002
for adding a discrete cosine transformed high resolution picture and a motion-compensated reference picture, and a frame memory
1003
for transient storage of the reference picture. The first downdecoder also includes a motion compensation unit
1004
for motion-compensating the reference picture stored in the frame memory
1003
with ½ pixel precision, and a downsampling unit
1005
for converting the reference picture stored in the frame memory
1003
to a picture of standard resolution.
This first downdecoder reduces an output picture, obtained on decoding as a high resolution picture by inverse discrete cosine transform, by the downsampling unit
1005
, to output resulting picture data with the standard resolution.
Referring to
FIG. 2
, the second downdecoder includes an inverse discrete cosine transform unit
1011
for performing 8×8 inverse discrete cosine transform, as it substitutes
0
for the high-frequency components of the discrete cosine transform (DCT) block of the high resolution picture, an adder
1012
for summing the discrete cosine transformed high resolution picture to the motion-compensated reference picture, and a frame memory
1013
for transient storage of the reference picture The second downdecoder also includes a motion compensation unit
1014
for motion-compensating the reference picture stored in the frame memory
1013
with ½ pixel precision, and a downsampling unit
1015
for converting the reference picture stored in the frame memory
1013
to a picture of standard resolution.
This second downdecoder performs inverse discrete cosine transform to obtain a decoded output picture, as a high-resolution picture, as it substitutes
0
for coefficients of high-frequency components among the totality of coefficients of the DCT block, and reduces the output picture in size by the downsampling unit
1015
to output picture data of standard resolution.
Referring to
FIG. 3
, a third downdecoder includes a decimating inverse discrete cosine transform unit
102
for doing e.g., 4×4 inverse discrete cosine transform, using only the coefficients of the low-frequency components of the DCT block of the bitstream of the high resolution picture, for decoding to a standard resolution picture, and an adder
1022
for summing the standard resolution picture processed with decimating inverse discrete cosine transform and the motion-compensated reference picture. The third downdecoder also includes a frame memory
1023
for transiently storing the reference picture and a motion compensation unit
1024
for motion-compensating the reference picture stored by the frame memory
1023
with a ¼ pixel precision.
In this third downdecoder, IDCT is executed using only low-frequency components of all coefficients of the DCT block to decode a picture of low resolution from a picture of high resolution.
The above-described first downdecoder performs inverse discrete cosine transform on the totality of the coefficients in the DCT block to obtain a high-resolution picture on decoding. Thus, the inverse discrete cosine transform unit
1001
of high processing capability and the frame memory
1003
of high capacity are needed. The second downdecoder performs discrete cosine transform on the coefficients in the DCT block to obtain a high-resolution picture on decoding, as it sets the high-frequency components of the coefficients to zero, so that a lower processing capacity of the inverse discrete cosine transform unit
1011
suffices. However, the fame memory
1003
of high capacity is yet needed. In contradistinction from these first and second downdecoders, the third downdecoder performs inverse discrete cosine transform on the totality of the coefficients in the DCT block, using only coefficients of the low-frequency components of the coefficients in the DCT block, so that a low processing capability of an inverse discrete cosine transform unit
1021
suffices. Moreover, since the reference picture of the standard resolution picture is decode, a lower capacity of the frame memory
1023
suffices.
Meanwhile, the display system of a moving picture in television broadcast is classified into a sequential scanning system and an interlaced scanning system. The sequential scanning system sequentially displays a picture obtained on sampling the totality of pictures in a given frame at the same timing. The interlaced scanning system alternately displays pictures obtained on sampling pixels in a given frame at different timings from one horizontal line to another.
In this interlaced scanning system, one of the pictures obtained on sampling pixels in a frame at different timings from one horizontal line to another is termed a top field or a first field, with the other picture being termed a bottom field or a second field. The picture containing the leading line in the horizontal direction of a frame becomes the top field, while the picture containing the second line in the horizontal direction of a frame becomes the bottom field. Thus, in the interlaced scanning system, a sole frame is made up of two fields.
With the MPEG2, not only a frame but also a field can be allocated to a picture as a picture compressing unit in order to compress the moving picture signals efficiently in the interlaced scanning system.
If, in the MPEG2, a field is allocated to a picture, the resulting bitstream structure is termed a field structure, while if a frame is allocated to a picture, the resulting bitstream structure is termed a frame structure. In the field structure, a DCT block is constituted by pixels in the field and discrete cosine transform is applied on the field basis. The
Goseki Masami
Kaneko Tetsuo
Komori Kenji
Mitsuhashi Satoshi
Sato Kazushi
Bugg George
Frommer William S.
Frommer & Lawrence & Haug LLP
Kelley Chris
Kessler Gordon
LandOfFree
Picture decoding method and apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Picture decoding method and apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Picture decoding method and apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3048032