2. Principle of H.264/AVC Normal Quantization Scheme

2.1. Scalar dead-zone quantization

In this section the principle of H.264/AVC normal quantization scheme is described in a generalized form.

A scalar quantizer for input signal W can be decomposed into a function Z=C[W] called a classification rule that selects an integer-valued class identifier called the quantization index at the encoder, and a reconstruction rule that produces a real-valued output W’=R[Z] at the decoder. Video encoder applies entropy coding to the quantization indices and communicates to the decoder. Although H.264/AVC JM reference software implements some classification functions, only reconstruction function is standardized.

In the quantization step of the encoder, the transform coefficients of the prediction error are quantized. This quantization is used to reduce the precision of the coefficients. Furthermore, the quantizer is designed to map insignificant coefficient values to zero whilst retaining a reduced [......]

Read More…

Permanent Link: Quantization Techniques in JM/KTA – Part 2

1. Overview

Currently most image and video coding systems and standards, such as MPEG-1/2 and H264/AVC, use transform-based techniques followed by quantization and entropy coding. The key idea is that transforms de-correlate the signal and compact the energy of a block into a few coefficients, which still represent the signal rather accurately after quantization and de-quantization. Nevertheless, this quantization/de-quantization process needs to be carefully designed in order to have the best possible subjective and objective quality.

In the encoder of H.264/AVC reference software, the scalar dead-zone quantization is adopted. In order to improve further the performance, other two adaptive quantization techniques are also introduced, which are both based on how to adjust the size of dead-zone and control the rounding behavior. In this tutorial, we will first introduce the principle of H.264/AVC normal quantization scheme, then discuss the adaptive rounding method which select adaptive[......]

Read More…

Permanent Link: Quantization Techniques in JM/KTA – Part 1

The 89^th MPEG / 31^st JVT / 38^th VCEG meetings are hosted by the Centre for Digital Music, School of Electronic Engineering and Computer Science of Queen Mary University of London, UK.

• AhGs will meet on 27 - 28 June 2009.
• The 89^th MPEG Meeting will be held on 29 June - 3 July, 2009.
• The 31^st JVT Meeting will be held on 29 June - 3 July, 2009.
• The 38^th ITU-VCEG Meeting will be held on 1 - 2 July, 2009. (Note: this VCEG meeting only discusses the future collaboration of VCEG and MPEG. Technical issues will be discussed in Geneva meeting on July 6 – 8.)

I will go to London to attend the meetings, and will report the progress in good time. Keep an eye on this website!

Permanent Link: The Upcoming 89th MPEG/31st JVT/38th VCEG Meetings

The technique of 1/8-pel interpolation [AD09] was proposed for motion-compensated prediction (MCP) and adopted in KTA software. Three types of interpolation filters are used for 1/2-, 1/4-, and 1/8-pel sub positions, respectively.

• [-3, 12, -39, 158, 158, -39, 12, -3]/256 for 1/2-pel sub positions.
• [-3, 12, -37, 229, 71, -21, 6, -1]/256 and [-1, 6, -21, 71, 229, -37, 12, -3]/256 for 1/4-pel sub positions.
• Bilinear filter for 1/8-pel sub positions.

  The frequency response of the interpolation filter is shown in the following figure. As can be seen, it is almost an ideal low-pass filter with a gain of 8 and a cutoff frequency π/8.

  According to the performance reported in the proposal, the gain on CIF/QCIF sequences is quite significant, i.e., up to 14% bit-rate reduction. I tested this technique based on a set of HD sequences. As shown in Table 1, the R-D performance is measured by BDPSNR [1], i.e., PSNR improvement at the same bit-rate or bit-rate reduction at the same PSNR.

 

 

 [......]

Read More…

Permanent Link: R-D Performance of 1/8-pel MCP on HD Sequences

In the Geneva meeting held in Feb. 2009, a proposal with the title “Video Coding Using Extended Block Sizes” was adopted by KTA, where the MB size is extended up to 64×64 and the motion partitions are scaled accordingly. At the same time, a 2D order-16 transform was also proposed for transforming the residual blocks with the size larger than or equal to 16×16. The transformation matrix of the proposed 2D order-16 transform is given as below, which is obtained by scaling the transformation matrix of 2D order-16 DCT by the factor 128 and rounding, and is non-orthogonal.

  Non-orthogonality will inevitably introduce transform error. Before analyzing the transform error quantitatively, let’s recall two properties of orthogonal transforms. Firstly, signals can be reconstructed perfectly if no quantization is performed in the transform domain. Secondly, if quantization is performed in the transform domain, the average variance (or energy) of the reconstruction er[......]

Read More…

Permanent Link: Transform Error Introduced by Non-orthogonality

The latest KTA software is JM11KTA2.3, which integrates the coding tools adopted in the Geneva meeting (Jan. 2009) and before.  

1. Inter prediction
   1. Adaptive interpolation filter (AIF)
      • 2-D non-separable AIF (AD08, AE16)
      • Separable AIF (COM16-C219, AG10)
      • Directional AIF (DAIF) (AG21, AG22, AH17, AH18)
      • Enhanced DAIF (E-DAIF) (AI12,  COM16-C126)
      • Enhanced DAIF 2 (E-DAIF2) (COM16-C125)
      • Enhanced AIF (EAIF) (COM16-C464, AI38, AJ30)
      • Switch interpolation filters with offsets (SIFO) (COM16-C463, AI35, AJ29, COM16-C126)
      • High precision filter (HPF) (AI33)
      • Single-Pass Encoding (AJ29, AK26)
   2. 1/8-pel MCP (AD09)
   3. Extended MCP block size (COM16-C123)
   4. Competition-based MV prediction (AC06r1)
2. Transform and quantization
   1. Mode-dependent directional transform (MDDT) (AF15, AG11, AH20, AJ24, AI36)
   2. Very large block transform (COM16-C123)
   3. Adaptive prediction error coding (APEC) (AB06, AD07, AE15)
   4. Adaptive quantization matrix selection (AQMS) (AC07, AD06, AF08, AI19)
   5. Rate-distortion optimized quantization (RDO[......]

Read More…

Permanent Link: KTA Software JM11KTA2.3

In the 37th VCEG meeting,  held in Yokohama on April 15-18, a potential “NGVC” project standing for next-generation video coding was discussed, which could mean either an extension of H.264/AVC or a new standard, depending on which form of standardization is determined to be appropriate for the technology design.

The preliminary requirements of the NGVC project focus on the following areas.

1. Coding efficiency:

• NGVC should be capable of providing a bit rate reduction of 50% at the same subjective quality

2. Complexity:

• NGVC should be capable of operating with a complexity ranging from 50% to 3 times H.264/AVC High Profile.
• When operated at a complexity of 50% compared to H.264/AVC High Profile, NGVC should provide a 25% bit rate savings compared to H.264/MPEG-4 AVC High Profile at equivalent subjective quality.

3. Applications:

• Low-delay interactive video communications
• Surveillance
• Streaming
• Broadcast
• Digital cinema and large-screen digital imagery
• Mobile video enter[......]

Read More…

Permanent Link: The Preliminary Requirements for NGVC

The VCEG and MPEG’s meeting plan for 2009 is listed as follows.

Jan. 27-Feb. 6 2009,    Geneva, CH              37th VCEG meeting + 30th JVT meeting 

Feb. 2-6 2009,             Lausanne, CH           87th MPEG meeting              

Apr. 15-18 2009          Yokohama, JP           37th VCEG meeting                    

Apr. 20-24 2009,         Maui, Hawaii, USA   88th MPEG meeting + 31st JVT meeting   

Jun. 29-Jul. 3 2009,     London, UK               89th MPEG meeting                                   

Oct. 20-30 2009,         Geneva, CH              38th VCEG meeting + 32nd JVT meeting  , 

Oct. 26-30 2009,         Xian, CN                    90th MPEG meeting

Permanent Link: VCEG and MPEG Meetings in 2009