Image analysis – Image compression or coding – Pyramid – hierarchy – or tree structure
Reexamination Certificate
1998-02-26
2001-09-25
Boudreau, Leo (Department: 2621)
Image analysis
Image compression or coding
Pyramid, hierarchy, or tree structure
C382S226000, C382S232000, C382S233000, C382S243000, C341S079000, C348S384100, C348S390100, C348S408100, C348S423100, C358S426010, C375S240000
Reexamination Certificate
active
06295380
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to object data processing apparatus, object data recording apparatus, data storage media, and data structure for transmission. More particularly, the invention relates to an apparatus for decoding compressed data, such as compressed digital video data, digital audio data, and program data, an apparatus for selecting desired data from the compressed data, an apparatus for recording the compressed data, a medium storing the compressed data, an apparatus for outputting the compressed data, and a data structure for transmitting the compressed data.
BACKGROUND OF THE INVENTION
In recent years, with the progress in information compression technology, a digital video/audio service providing video information and audio information by digital signals has been put to practical use for broadcasting media, such as ground broadcasting, satellite broadcasting, and CATV.
Under the existing circumstances, as a compressive coding method for the next generation, an object coding method has attracted attention. This object coding method is not to uniformly compress the whole image, i.e., video data corresponding to a single image, but to compress video data corresponding to d single image in units of individual objects constituting the image while paying attention to the contents of the image.
When video data corresponding to a single image is subjected to the compressive coding in object units, compressed (coded) video data is separable corresponding to the respective objects, whereby a specific object in the image can be extracted or replaced.
Meanwhile, as a method of implementing a data transmission format for making the best use of the object coding method, a method of multiplexing compressed video data, audio data, and other digital data is discussed.
There is MPEG4 as an international standard of a method of multiplexing data compressed by the object coding method (ISO/IEC JTC1/SC29WG11 N1483, “System Working Draft”, November 1996). Hereinafter, a description is given of the data multiplexing method based on MPEG 4 and a method for reproducing the multiplexed data, with reference to figures.
FIG. 18
is a diagram for explaining the object coding method. In the figure, reference numeral
120
designates a scene (an image) in a series of images obtained from video data with audio. This scene
120
is composed of a plurality of objects (sub-images) making a hierarchical structure. To be specific, the scene
120
is composed of three objects: a background image (background)
121
, a moving object
122
that moves in the background, and a background audio
123
attendant on the c background. The moving object
122
is composed of four objects: a first wheel
124
, a second wheel
125
, a body
126
, and a moving object audio
127
attendant on the moving object. Further, the object of body
126
is composed of two objects: a window
128
and the other part
129
. In the hierarchical structure, the objects
121
~
123
belong to the uppermost first layer L
1
, the objects
124
~
127
belong to the second layer L
2
lower than the first layer L
1
, and the objects
128
and
129
belong to the third layer L
3
lower than the second layer L
2
.
In the object coding method, scene data corresponding to the scene
120
are compressively (coded in units of the lowermost objects constituting the scene
120
. In other words, scene data corresponding to the scene
120
are compressively coded for each of the objects
121
,
123
,
124
,
125
,
127
,
128
and
129
.
FIG. 19
is a diagram for explaining a data structure for transmitting coded data corresponding to the respective objects mentioned above, which is obtained by performing object coding to the scene data of the scene
120
.
In
FIG. 19
, MEg shows a multiplexed bit stream having a prescribed format, obtained by multiplexing coded data of the respective objects and auxiliary data. This multiplexed bit stream MEg is transmitted as coded data corresponding to the scene data.
The multiplexed bit stream MEg is partitioned into plural packets in prescribed units, i.e., each packet having prescribed number of bytes, and coded data of the respective objects are allocated to the packets having their own values (SLC=1, 2, . . . ) as logical channels (LC).
To be specific, in the multiplexed bit stream MEg shown in
FIG. 19
, coded video data of object [1] is allocated to packets Pa
3
and Pa
6
having a logical channel SLC=3, coded video data of object [2] is allocated to packets Pa
5
and Pa
7
having a logical channel SLC=4, and coded audio data of object [3] is allocated to a packet
4
having a logical channel SLC=5. information relating to the byte number of packet when multiplexed, the logical channel LC of each packet, and the packet transmission order is allocated as control information to a packet having another logical channel (not shown) for transmission.
The objects [1] and [2] are the background image
121
and the moving object
122
shown in
FIG. 18
, respectively, and the object [3] is the background audio
123
shown in FIG.
18
.
In the multiplexed bit stream MEg, allocated to the packet Pal of Logical channel SLC=1 is information relating to a scene composition method for regenerating the scene composed of the respective objects (composition stream), and allocated to the packet Pa
2
of logical channel SLC=2 is information showing how the coded data of the respective objects are multiplexed (stream association table).
Accordingly, when a plurality of coded data obtained by object coding are multiplexed and transmitted, with the coded data of the respective objects, the composition stream showing the structure of a scene composed of the objects and the stream association table showing the correlation of the transmitted streams (each stream being a series of coded data corresponding to each object) are transmitted simultaneously.
FIG. 20
is a diagram for explaining a scene description according to the composition stream, illustrating a description SD corresponding to the single image (scene)
120
shown in FIG.
18
.
In the scene description SD according to the composition stream, the image
120
is shown by Scene
140
, and the fact that the image
120
shown by Scene
140
is composed of the background image
121
, the moving object
122
, and the background audio
123
is shown by Video(1)
141
, Node(1)
142
, and Audio(1)
143
, respectively. Here, Scene
140
, Video(1)
141
, Node(1)
142
, and Audio(1)
143
are descriptors describing the image
120
, the background image
121
, the moving object
122
, and the background audio
123
shown in
FIG. 18
, respectively.
Further, in the scene description SD, the fact that the moving object
122
shown by Node(1)
142
is composed of the first wheel
124
, the second wheel
125
, the body
126
, and the moving object audio
127
is shown by Video(2)
144
, Video(3)
145
, Node(2)
146
, and Audio(2)
147
, respectively, which are descriptors corresponding to these objects.
Further, the fact that the body
126
shown by Node(2)
146
is composed of the window
128
and the other part
129
is shown by Video(4)
148
and Video(5)
149
, respectively, which are descriptors corresponding to these objects.
Each of the descriptors is given a stream index (stream id) for identifying a stream corresponding to coded data of each object in the multiplexed bit stream MEg. To be specific, as shown in
FIG. 20
, stream indices Sid=1~Sid=5 are given to the descriptors
141
~
145
, respectively, and stream indices Sid=6, Sid=7, and Sid=8 are given to the descriptors
148
,
149
, and
147
, respectively. Sid is a specific number of each stream id.
Accordingly, it can be seen from the scene description SD according to the composition stream that a scene is composed of what kinds of objects. However, the scene description SD according to the composition stream does not describe how the coded data corresponding t
Boudreau Leo
Mariam Daniel G.
Matsushita Electric - Industrial Co., Ltd.
Parkhurst & Wendel L.L.P.
LandOfFree
Object data processing apparatus, object data recording... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Object data processing apparatus, object data recording..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Object data processing apparatus, object data recording... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2446575