Game Development Reference
In-Depth Information
Table 8.1 Classification of the generalized MBC
Model-based coding
Example of used models
Example of coding schemes
Pixel-based coding
Markov signal model Gaussian
model
PCM
Block-based coding
Block-based translation model
MPEG-x H.26x series of
standards
Geometric partition coding
Triangle, mesh-based model
Mesh-based coding
Segmentation-based coding Region-based motion model
Contour coding
Object-based coding
Object-based motion model
MPEG-4 object-based coding
Knowledge-based coding
Facial model
Facial model-based image
coding
Semantic-based coding
Facial expression model
Model-based facial expression
coding
Perceptual coding
HVS model
Texture analysis and synthesis
Intelligent coding
Composite signal perceptual
model
Learning-based coding, cloud
coding, layered coding,
surveillance coding
with block-based motion model coding later. Simultaneously with the development
of block model-based coding, other kinds of models are also developed.
Firstly, as a variation of block-based motion model, more flexible geometric
partitions were proposed for motion compensation to represent the rotation, zoom
and other motions, including deformable block motion compensation (Seferidis and
Ghanbari 1993 ), control grid or mesh (Sullivan and Baker 1991 ), and other geometric
partitions. In Kubasov and Guillemot ( 2006a ), a codec which supports both the block
and mesh-based motion compensation mode was proposed. Recently, mesh-based
coding was also studied for Distributed Video Coding (DVC) and 3DV video coding
(Kubasov and Guillemot 2006b ; Smolic et al. 2007 ).
As a further evolution of geometric partition coding, contour codingwas proposed,
which can be viewed as a predecessor of segmentation coding. In geometric partition
coding, the image content is usually partitioned into small geometric pieces to model
themotionmore precisely. However, for segmentation coding, the contours or regions
of the segmentation can provide more compact representations to the image content
and hence can achieve higher compression performance. For the segmentation-based
coding, the key issue is how to perform image segmentation efficiently, which gives
a content-based representation of the image. With the advance on contour coding
and image segmentation (Horowitz and Pavlidis 1976 ; Beaulieu and Goldberg 1989 ;
Kocher and Leonardi 1986 ; Haralick and Shapiro 1985 ), the segmentation-based
coding was emerging as a hot topic in the 1980s. Based on image segmentation,
a segmented image coding scheme was proposed in Biggar et al. ( 1988 ), and the
scheme shows significant performance improvement compared to the traditional
DCT transform-based coding. In Horowitz and Pavlidis ( 1976 ), a generalized shape-
adapted transform coder was proposed by developing a set of basis functions which
are orthogonal with respect to the shape to be code segmented. Furthermore, these
 
 
Search WWH ::




Custom Search