Coding efficiency/Compression ratio: The loss of information or distortion measure:

VI. Video Compression

Coding efficiency/Compression ratio:

6.1 Basci concepts and Tech. of Video Coding-Introduction

The loss of information or distortion measure:

6.1 Basci concepts and Tech.s of Video Coding-Introduction

The functional elements of the generalized model:

Spatial operator Quantizer Variable length coding

The elements required in the feedback loop: Inverse operators Delayed frame memory Motion estimation Motion compensation

6.1 Basci concepts and Tech. of Video Coding-Introduction

A fundamental issue facing the design of video compression systems: the tradeoff between bitrate and quality or distortion

Factors in design and selection of a video compression system:

Video characteristic Transmission requirement Compression system characteristics and

performance Rate-distortion requirement Standards requirement

6.1 Basci concepts and Tech. of Video Coding-Application Requirements

6.1 Basci concepts and Tech.s of Video Coding-Application Requirements

Sampling of Analog Video Signals Two-dimensional spatial sampling and one-

dimensional temporal sampling Nyquist sampling theorem: defines the

conditions under which sampled analog signals can be perfectly reconstructed.

6.1 Basci concepts and Tech. of Video Coding-Digital Video Signals and Formats

Digital Video Formats

6.1 Basci concepts and Tech.s of Video Coding-Digital Video Signals and Formats

Entropy and Predictive Coding Discrete memoryless source (DMS) – VLC

coding: Huffman and arithmetic coding Markov source – predictive coding: differential pulse code modulation coding

(DPCM)

6.1 Basci concepts and Tech.s of Video Coding-Video Compression Tech.s

Block Transform Coding: The Discrete Cosine Transform

Has fast implementations using real calculations

Does not produce any significant discontinuities at the block edges when reconstruction


Block Transform Coding: The Discrete Cosine Transform

Has fast implementations using real calculations

Does not produce any significant discontinuities at the block edges when reconstruction


Quantization Uniform scalar quantizer Non-uniform quantization Vector quantizers


Motion Compensation and Estimation Motion estimation and compensation are

common techniques used to encode the temporal aspect of a video signal.

Three stages of motion compensated video coding:

1. motion estimation 2. motion compensation 3. encode the prediction error


Motion Compensation and Estimation Motion estimation is an interframe prediction

process falling in two general categories: 1. pel-recursive algorithms 2. block-matching algorithms: measure and

search


The H.261 recommendation is targeted at the videophone and videoconferencing application market running on connection-based ISDN at p x 64 kbps, p = 1, . . . ,30.

It explicitly defines the encoded bit stream syntax and decoder, while leaving the encoder design to be compatible with the decoder specification.

The video encoder is required to carry a delay of less than 150 ms so that it can operate in real-time bidirectional videoconferencing applications.

6.1 Basci concepts and Tech.s of Video Coding-H.261 Standards

The H.261 is part of a group of related ITU recommendations that define visual telephony systems.

This group includes the following:





Motion Compensation Transformation and Quantization Scalabilities --one scalable coded file that offers

increasingly greater spatial resolution, higher frame rates, or a better signal-to-noise ratio.

6.2 Spatiotemporal Wavelet Video Compression-Video Compression Basics

Hybrid Wavelet Coder temporal differential PCM (DPCM)+wavelet

based coders in the spatial dimension

Spatiotemporal Wavelet Coder Without motion compensation With motion compensation

Zero Coding and Embedding

6.2 Spatiotemporal Wavelet Video Compression-Wavelet Compression

Segmenting objects Temporal linking of objects Encoding objects “Inter” mode—encoding objects by the

motion parameters (I objects) “Intra” mode—encoding those that can

not be predicted (P objects)

6.2 Spatiotemporal Wavelet Video Compression-Object-Based Wavelet Compression

Joint Motion Estimation and Segmentation


The likelihood functional that describes how well the observed images match the motion field data:



A priori density of motion and enforces prior constraints on the motion field:



A priori expectations for the object label field itself:

Maximation ApproachMAP solutionTwo-step iterative hierarchical

procedure: motion estimation & segmentation


Coding of Video Objects I mode: motion compensated predictive

(MCP) coding in hybrid object-based (OB)-MCP

P mode: spatial wavelet coding


Object Motion/Segmentation Coding Object-based 3-D wavelet coding (OB-

3DSBC) coder



Represent visual data in terms of regions, defined by their contour and texture, possibly corresponding to objects or to parts of objects.

emphasize visually sensitive data while neglecting visually insignificant information

Second-generation concept is now widely accepted and has become the basic philosophy of the new MPEG-4 standard

Is called dynamic coding

6.3 Object-Based Video Coding-Introduction

6.3 Object-Based Video Coding-Second-Generation Coding

6.3 Object-Based Video Coding-Object-Based Video Coding

Intramode(I)—an object is coded independently

Intermode(P)—a video object is coded taking into account information available on its past


Object Shape and Geometry Coding A progressive contour coding based on a

polygonal approximation of the shape boundary is used to code the outline of object.

Suitable for sketch-based retrieval that is based on video object shapes

A geometrical shape boundary description can be integrated into an object-based mesh coding scheme



6.3 Object-Based Video Coding-Object-Based Video Coding Lossless shape representation-- altered

boundary triangles


Object Motion Estimation, Compensation, and Coding

Texture Representation


Coding efficiency/Compression ratio: The loss of information or distortion measure:

Documents

s of video codingintroduction

video compression vi

basci concepts

formatsdigital video

s block transform coding

smotion compensation

estimationmotion estimation

block edges