2006•Unpublished venueRequires access

A Fast Mode Selection Algorithm in H.264 Video Coding

Dong‐Hyung Kim, Jechang Jeong

Open publisher page 13 citations

Abstract

For improvement of coding efficiency, the H.264 video coding standard uses new coding tools, such as variable block size, quarter-pixel-accuracy motion estimation, multiple reference frames, intra prediction, loop filter, etc. Using these coding tools, H.264 achieves significant improvement in coding efficiency compared with existing standards. However, encoder complexity increases tremendously. Among the tools, the macroblock mode selection and the motion estimation contribute most to total encoder complexity. This paper focuses on complexity reduction in macroblock mode selection. Of all macroblock modes which can be selected, inter8×8 and intra4×4 have the highest complexity. We propose two methods for complexity reduction of inter8×8 and intra4×4 by using the costs of the other macroblock modes. Simulation results show that the proposed methods save up to 57.7% of total encoding time compared with the H.264 reference implementation, whereas the average PSNR only decreases less than 0.05dB.

About this research paper

What this paper is about

For improvement of coding efficiency, the H.264 video coding standard uses new coding tools, such as variable block size, quarter-pixel-accuracy motion estimation, multiple reference frames, intra prediction, loop filter, etc. Using these coding tools, H.264 achieves significant improvement in coding efficiency compared with existing standards. However, encoder complexity increases tremendously. Among the tools, the macroblock mode selection and the motion estimation contribute most to total encoder complexity. This paper focuses on complexity reduction in macroblock mode selection. Of all macroblock modes which can be selected, inter8×8 and intra4×4 have the highest complexity. We propose two methods for complexity reduction of inter8×8 and intra4×4 by using the costs of the other macroblock modes. Simulation results show that the proposed methods save up to 57.7% of total encoding time compared with the H.264 reference implementation, whereas the average PSNR only decreases less than 0.05dB.

Why it matters

OpenAlex reports 13 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

For improvement of coding efficiency, the H.264 video coding standard uses new coding tools, such as variable block size, quarter-pixel-accuracy motion estimation, multiple reference frames, intra prediction, loop filter, etc. Using these coding tools, H.264 achieves significant improvement in coding efficiency compared with existing standards. However, encoder complexity increases tremendously. Among the tools, the macroblock mode selection and the motion estimation contribute most to total encoder complexity. This paper focuses on complexity reduction in macroblock mode selection. Of all macroblock modes which can be selected, inter8×8 and intra4×4 have the highest complexity. We propose two methods for complexity reduction of inter8×8 and intra4×4 by using the costs of the other macroblock modes. Simulation results show that the proposed methods save up to 57.7% of total encoding time compared with the H.264 reference implementation, whereas the average PSNR only decreases less than 0.05dB.

Key concepts: Macroblock, Computer science, Coding tree unit, Encoder, Coding (social sciences), Algorithmic efficiency, Algorithm, Motion estimation

Related papers

Back to paper searchBrowse research topicsOriginal source
A Fast Mode Selection Algorithm in H.264 Video Coding — Research Paper | ScholarLens