Image analysis – Image compression or coding – Including details of decompression
Patent
1998-03-27
2000-06-13
Mehta, Bhavesh
Image analysis
Image compression or coding
Including details of decompression
382305, G06K 946
Patent
active
060758996
DESCRIPTION:
BRIEF SUMMARY
FIELD OF THE INVENTION
The present invention relates to an image decoding apparatus used for decoding both video and still images and to an image memory.
PRIOR ART
There is an ever-expanding demand for video decoding apparatuses which are used by a variety of multimedia appliances to decode compressed video data. Such compressed video data is produced by encoding video into digital data where only the differences between successive frames are included to reduce the total data size.
When decoding compressed video data, the images to be displayed are produced by combining images used as a reference for differential calculation (hereinafter, "reference pictures") with coded images (hereinafter, "differential pictures") for calculated differences between images.
MPEG (Motion Pictures Experts Group) methods use images called I-pictures ("Intra Pictures") as reference pictures. Such I-pictures are compressed image data, but are characterized by including luminance component data and chrominance component data for an entire frame. MPEG methods use P-pictures (Predictive Pictures) and B-pictures (Bidirectionally Predictive Pictures) as differential images. Of these, P-pictures are image data for calculated differences with a preceding frame, while B-pictures are image data for calculated differences with both a preceding and a succeeding frame.
B-pictures and P-pictures are generated so that the differences between frames are calculated in units called pixel blocks that are each composed of 16 horizontal pixels by 16 vertical pixels. Each pixel block is composed of a luminance block comprising 16*16 luminance components as shown in FIG. 1A and, as shown in FIG. 1B, a blue chrominance block ("Cb block") composed of 8*8 blue chrominance components and a red chrominance block ("Cr block") composed of 8*8 red chrominance components. The luminance block of 16*16 luminance components, and the red and blue chrominance blocks of 8*8 chrominance components are collectively called a "macro block", with this being a unit for image decoding.
Each macro block is encoded as follows. A discrete cosine transform ("DCT") is performed for the luminance block and the red and blue chrominance blocks and the spatial frequency components in the low frequency band are separated from the spatial frequency components in the high frequency band. The high frequency components are discarded and quantization is performed for the spatial frequency components located in the low frequency band.
When decoding an encoded macro block, inverse quantization and an inverse DCT are first performed, before motion compensation is performed based on motion vectors so that images can then be displayed. Motion vectors are information that indicates areas with the highest correlation based on a comparison of the preceding and succeeding frames. This is to say, motion vectors are information that shows how objects such as people move within the picture with respect to the preceding and succeeding frames, in units of blocks.
The motion compensation referred to here means the addition (blending) of the appropriate differences to the reference pictures displayed before or after the present frame to form one frame that is a display image.
The compression and decompression techniques for video images under MPEG standard which are referred to by the present application are as described above. These are explained in more detail in a variety of publications, such as the Japanese publication "Pointo Zukaishiki Saishin MPEG Kyokasho (Latest MPEG techniques--An Illustrated Reader) by ASCII Publishing Inc.
When decoding video data that has been compressed by coding the differences between frames, it is necessary to use a DRAM (Dynamic Random Access Memory) or an SDRAM (Synchronous Dynamic Random Access Memory). This is because it is necessary to accumulate a frame image to which the calculated differences are to be added when decoding the compressed video data. With its low cost and large capacity, SDRAM is well suited to this accumulation of images.
While a video de
REFERENCES:
patent: 5905821 (1999-05-01), Nonoshita et al.
patent: 5960115 (1999-09-01), Kao et al.
patent: 5963678 (1999-05-01), Nozawa
Hirai Makoto
Kimura Kozo
Kiyohara Tokuzo
Yoshioka Kosuke
Matsushita Electric - Industrial Co., Ltd.
Mehta Bhavesh
LandOfFree
Image decoder and image memory overcoming various kinds of delay does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image decoder and image memory overcoming various kinds of delay, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image decoder and image memory overcoming various kinds of delay will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2076217