Computer graphics processing and selective visual display system – Computer graphics display memory system – Memory allocation
Reexamination Certificate
1999-02-12
2002-10-08
Tung, Kee M. (Department: 2671)
Computer graphics processing and selective visual display system
Computer graphics display memory system
Memory allocation
C345S555000, C348S445000, C348S512000, C348S714000, C375S240250, C375S240260, C375S240010
Reexamination Certificate
active
06462744
ABSTRACT:
This application is based on application No. H10-030963 filed in Japan, the content of which is hereby incorporated by reference.
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image decoding apparatus, and more specifically relates to a technique for improving the utilization ratio of a memory.
2. Description of the Background Art
MPEG2 (Moving Pictures Experts Group2) standard is highly regarded by technicians all over the world as the international standard technique for compressing moving pictures. In particular, when encoded according to an intra-frame bidirectionally predictive method, moving pictures are compressed at a. very high ratio using interrelations between a picture and its previously and subsequently displayed pictures. This method is of particular value when recording moving pictures onto a removable storage medium or transmitting moving pictures to a remote destination via a communication medium.
While moving pictures are compressed at a high ratio with this method, there is apprehension among technicians that image decoding apparatuses may increase in complexity and scale. This is because when decoding picture data that is encoded according to the intra-frame bidirectionally predictive method (such picture data being ordinarily referred to as “B-pictures (Bidirectionally Predictive pictures)”), an image decoding apparatus needs to refer to previously and subsequently displayed pictures. This means that the frame memories of the image decoding apparatus need to store both previously displayed pictures and subsequently displayed pictures that are referred to by B-pictures. As a result, the scale of the frame memories is increased. Each frame memory is a memory area for storing pixel data that is displayed on a screen for one frame. Pixel data stored in the frame memory is read and is converted into image signals in line units (720 horizontal pixels by one vertical pixel, for instance) in synchronization with a horizontal synch signal of the display.
FIG. 1A
shows an example of an SD-RAM (Synchronous Dynamic RAM) where three frame memories are reserved. This example is based on the assumption that the SD-RAM is two banks by 2048 rows by 256 columns in size. There are pictures of three picture types, which is to say the aforementioned B-pictures, Intra pictures (I-pictures), and Predictive pictures (P-pictures). I-pictures are encoded according to an intra-frame encoding method and B-pictures are encoded according to an intra-frame forward direction predictive encoding method. The SD-RAM shown in
FIG. 1A
includes three frame memories which store decoded pictures according to the picture types.
More specifically, in this drawing, the non-reference picture data frame memory stores a decoded B-picture; the reference picture data A frame memory stores a decoded I-picture or a decoded P-picture; and the reference picture data B frame memory stores a decoded I-picture or a decoded P-picture. Each frame memory occupies an area which is two banks by 608 rows by 256 columns in size in the SD-RAM. Decoded picture data stored in the reference picture data A frame memory and the reference picture data B frame memory is referred to while a B-picture is decoded.
A large area in the SD-RAM is allocated to these frame memories under the stated conventional technique. Therefore, there is a problem that the SD-RAM cannot provide work areas aside from the frame memories. Work areas are used for several purposes, such as for storing on-screen display (OSD) data. The word “OSD” refers to character fonts and computer graphics overlaid on moving pictures. The OSD data is used to display the operation states of an image decoding apparatus, such as “play”, “stop”, and “record”, on a screen. The OSD data is also used to display a menu for allowing an operator to input desired instructions.
When the SD-RAM does not include enough space for storing OSD data, an additional memory needs to be used to provide an OSD data storage area for storing the OSD data. When such an additional memory cannot be used, a memory mapping method may be used to provide the OSD data storage area, as shown in FIG.
1
B. In
FIG. 1A
, the non-reference picture data frame memory occupies a memory area which is 2 banks by 608 rows by 256 columns in size. On the other hand, in
FIG. 1B
, the memory area assigned to the non-reference picture data frame memory is reduced to the size of 2 banks by 507 rows by 256 columns. As a result, a free area which is 2 banks by 101 rows by 256 columns in size is obtained and the free area is used as the OSD data storage area.
This technique where the data size of B-pictures is reduced is based on the characteristic that B-pictures are not referred to by other pictures. While degradation in the picture quality of an I-picture or P-picture stored in the frame memories may adversely affect the picture quality of other pictures that refer to the I-picture or P-picture, a B-picture stored in a frame memory is not referred to by other pictures. Therefore, the degradation in the picture quality of B-pictures does not adversely affect the picture quality of other pictures. As a result, a reduction in the size of the frame memory area for storing picture data of a B-picture is admissible, unlike the cases of an I-picture or a P-picture. Accordingly, the frame memory for storing picture data of B-pictures is made smaller than the frame memories for storing picture data of I-pictures and P-pictures.
To reduce the size of picture data of B-pictures, the chrominance data of B-pictures is generated at an interval (is subsampled). This method is based on the human eye's sensitivity to small changes in brightness and relative insensitivity to small changes in color.
However, subsampling of the chrominance data of B-pictures accentuates the differences in picture quality between the display of I-pictures and P-pictures and the display of B-pictures. That is, picture quality is not degraded when I-pictures and P-pictures are displayed, but picture quality is degraded when B-pictures are displayed. Therefore, when moving pictures are replayed, pictures of degraded picture quality and pictures of non-degraded picture quality are alternately displayed on the screen. Furthermore, when recording moving pictures onto a removable storage medium or transmitting moving pictures to a remote destination via a communication medium, the moving pictures need to be compressed at a high ratio. Therefore, many frames are encoded as B-pictures and few frames are encoded as I-pictures and P-pictures. Accordingly, when the picture quality of the frequently occurring B-pictures is degraded, an operator may have the impression that the picture quality of the moving pictures as a whole is poor.
A playback apparatus ordinarily requires an OSD data storage area when an OSD needs to be displayed according to the instruction input by an operator. Since an OSD is displayed according to an operator's instruction, the OSD data storage area only needs to be reserved when OSD data is displayed. However, a work area is always reserved under the stated conventional technique, which causes a problem that the picture quality of B-pictures is sacrificed unnecessarily.
SUMMARY OF THE INVENTION
The object of the present invention is to provide an image decoding apparatus which reserves a work area for storing various data, such as OSD data, without sacrificing picture quality.
To achieve the state object, the image decoding apparatus of the present invention decodes pieces of picture data included in a video stream according to an instruction from a host apparatus and writes the decoded pieces of picture data into a storage apparatus, where each piece of picture data is of one of a plurality of picture types which are each related to one of different encoding methods, and the storage apparatus includes a plurality of frame areas for storing most recently decoded pieces of picture data separately according to the plurality of picture types, where the image decoding apparatus inclu
Hirai Makoto
Kiyohara Tokuzo
Mochida Tetsuji
Nishida Hideshi
LandOfFree
IMAGE DECODING APPARATUS THAT PERFORMS IMAGE DECODING SO... 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 DECODING APPARATUS THAT PERFORMS IMAGE DECODING SO..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and IMAGE DECODING APPARATUS THAT PERFORMS IMAGE DECODING SO... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2984289