Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
1997-09-25
2001-10-23
Le, Vu (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
C375S240240
Reexamination Certificate
active
06307885
ABSTRACT:
BACKGROUND OF THE INVENTION
1 Field of the Invention
The present invention relates to a device for and method of coding/decoding image information by which the image transferred from an image input device is divided into object and background images having predetermined shape information, prior to an encoding/decoding and, more particularly, to a device for and method of coding/decoding image information which is derived to enhance coding/decoding and transmission efficiencies when coding image information by merging a plurality of boundary blocks, by performing a variable length coding according to the characteristic of each block transformed after the merge, and/or decoding the coded information.
2. Discussion of Related Art
As is generally known, image processing techniques, rather than using an entire comprehensive coding, typically divide an image of one frame into designated unit blocks having predetermined shape information, and subsequently the unit blocks are each processed by a compressive coding.
When a still picture is entered, the image is divided into object and background images for a later transmission. As for a moving picture, the variations of the object image are first transferred. By this process, natural or artificial images are composed and decomposed in the units of an object image unrestrictedly, enhancing the compressive coding and transmission efficiencies. An international standard that is based on the unit blocks having shape information has been established by the international organization for standardization (hereinafter, referred to as “ISO”), international telecommunication union telecommunication standardization sector (hereinafter, referred to as “ITU-T”) and the like.
For example, ISO/IEC affiliated organizations are carrying out projects for a moving picture compression standardization; MPEG(Moving Picture Expert Group)-4 for moving picture compression standardization in WG11, JPEG(Joint Photographic Coding Experts Group)-2000 for a still picture compression standardization, and H.263+, H.320 and H.331 in ITU-T.
MPWG-4 is based on the concept of shape information and will be described below.
The concept of a VOP (Video Object Plane) is used in the MPEG-4 as a unit block having designated shaped information.
The VOP is defined as a tetragon that includes object and background images divided from an input picture.
The keynote of MPEG-4 lies in the fact that when a picture has a designated object or object region, the object image is divided into VOPs, each of which will be separately encoded, transmitted, or decoded.
The concept of a VOP is used in processing object images in the field of computer graphics and multimedia such as Internet multimedia, interactive video games, interpersonal communications, interactive storage media, multimedia mailing, wireless multimedia, networked database services using an ATM (Asynchronous Transfer Mode) network and the like, remote emergency systems, and remote video surveillance.
FIG. 1
is a block diagram of the VM (Verification Model) encoder 100 first decided by international standardization affiliated organization (ISO/IEC JTC1/SC29/WG11 MPEG96/N1172 JANUARY).
As shown in
FIG. 1
, a VOP definition block
110
divides a picture sequence, to be transmitted or stored, into a unit object image, and defines different VOPs.
FIG. 2
shows a VOP having a “cat” picture as an object image.
As shown in
FIG. 2
, the horizontal size of the VOP is defined as “VOP width”, and the vertical size is “VOP height”. Thus the defined VOP is then divided into (M×N) macro blocks consisting of M and N pixels along the X and Y axes. A grid starting point is framed at the left top of the VOP. For example, the VOP is divided into (16×16) macro blocks having 16 pixels along the X and Y axes respectively.
If the macro blocks formed in the right and bottom part of the VOP do not have M and N picture elements each along the X and Y axes, the VOP should be extended in size to contain M and N pixels respectively along the X and Y axes.
Both M and N are determined as even numerals so that an encoding can be performed in a texture coding sub block, as is described below.
FIGS. 3
a-
3
b
illustrates a VOP formed by extracting an object image (having a designated shape) from an input picture, and divided into unit macro blocks.
As shown in
FIGS. 3
a-
3
b
, the macro blocks forming the VOP comprises regions with object image information and ones having no object image information.
Referring to
FIG. 3
a
, the respective macro blocks are divided into interior macro blocks having object image information exterior macro blocks having no object image information, and boundary macro blocks partly including the image information. Prior to coding or decoding, the macro blocks are divided into the above-mentioned classes.
Referring to
FIG. 3
b
, before a coding or decoding is performed, the boundary macro blocks are divided into, interior sub blocks having object image information, exterior sub blocks having no object image information, and boundary sub blocks partly having object image information.
The respective VOPs defined by the VOP definition block
110
are transferred into VOP coding blocks
120
a
,
120
b
, . . . , and
120
n
to perform a coding by VOPs. They are then multiplexed in a multiplexer
130
and transmitted as bit streams.
FIG. 4
is a block diagram of the VOP coding blocks
120
a
,
120
b
, . . . , and
120
n
of the VM encoder
100
as decided by international standardization affiliated organizations.
Referring to
FIG. 4
, a motion estimation block
121
receives the VOP concerning the respective object images in order to estimate motion information in the macro blocks from the VOP received.
The motion information estimated by the motion estimation block
121
is transferred into a motion compensation block
122
.
An adder
123
receives the VOP, whose motion is compensated by the motion compensation block
122
, and the value detected by the adder
123
is transferred into a texture coding block
124
for encoding texture information of the object as sub blocks.
For example, each of the (16×16) macro blocks is divided into (8×8) sub blocks comprising (M/2×N/2) pixels each along the X and Y axes of the macro block.
An adder
125
obtains the sum of the VOP motion-compensated by the motion compensation block
122
and the texture information encoded by the texture coding block
124
. The output of the adder
126
is transferred into a previous reconstructed VOP block
126
for detecting the previous VOP, which is the VOP of the previous image.
The previous VOP detected by the previous reconstructed VOP block
126
is used in the motion estimation block
121
and the motion compensation block
122
so as to estimate and compensate the motion.
The VOP defined by the VOP definition block
110
is transferred into a shape coding block
127
for coding the shape information.
As indicated by dotted lines, the output of the shape coding block
127
is selectively transferred into the motion estimation block
121
, the motion compensation block
122
, or the texture coding block
124
for the use purpose in motion-estimating, motion-compensating, or encoding the texture information of an object. This is determined by the application field of the VOP coding blocks
120
a
,
120
b
, . . . , and
120
n.
The motion information estimated by the motion compensation block
121
, the texture information encoded by the texture coding block
124
, and the shape information encoded by the shape coding block
127
are multiplexed by a multiplexer
128
, and they are transmitted as a bit stream into the multiplexer
130
as shown in FIG.
1
.
As shown in
FIG. 5
, a demultiplexer
210
of a VM decoder
200
divides the VOP signal encoded by the VM encoder
100
into VOPs. The respective VOP signals divided by the demultiplexer
210
are decoded into the original VOP picture by a plurality of VOP decoding Blocks
220
a
,
220
b
, . . . , and
220
n
, and composed by a composition block
230
.
FIG. 6
is a b
Kweon Ji-Heon
Moon Joo-Hee
Bryan Cave LLP
Hyundai Electronics Ind. Co. Ltd.
Kurland, Esq. Lawrence G.
Le Vu
LandOfFree
Device for and method of coding/decoding image information does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Device for and method of coding/decoding image information, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Device for and method of coding/decoding image information will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2587377