• Title/Summary/Keyword: Exp-Golomb VLC

Search Result 4, Processing Time 0.018 seconds

An Efficient Dynamic Entropy Coding by using Multiple Codeword in H.264/AVC (다중 부호어를 이용한 효율적인 H .264/AVC 동적 부호화 방법)

  • 백성학;문용호;김재호
    • The Journal of Korean Institute of Communications and Information Sciences
    • /
    • v.29 no.8C
    • /
    • pp.1055-1061
    • /
    • 2004
  • In this paper, we propose an efficient dynamic coding scheme by using multiple codewords in H.264/AVC entropy coding. The exponential Golomb (Exp-Golomb) codewords used in H.264/AVC do not reflect enough the symbol distributions of the combined syntax element in (7) due to their static probability distribution characteristics. However, the multiple codewords in this paper have different statistical characteristics. we propose a dynamic coding scheme by using selectively among multiple codewords to encode the combined syntax element according to given image sequences. Simulation results show that our proposed scheme outperforms the existing (7) method in compression efficiency without loss of quality.

An Efficient Dynamic Entropy Coding by using Multiple Codeword in H.264/AVC (다중 부호어를 이용한 효율적인 H.264/AVC 동적 부호화 방법)

  • Baek, Seong-hak;Moon, Yong-hoo;Kim, Jae-ho
    • The Journal of Korean Institute of Communications and Information Sciences
    • /
    • v.29 no.9C
    • /
    • pp.1244-1251
    • /
    • 2004
  • In this paper, we propose an efficient dynamic coding scheme by using multiple code words in H.264/A VC entropy coding. The exponential Golomb (Exp-Golomb) code words used in H 264/A VC do not reflect enough the symbol distributions of the combined syntax element in [7] due to their static probability distribution characteristics However, the multiple codewords in this paper have different statistical characteristics. we propose a dynamic coding scheme by using selectively among multiple codewords to encode the combined syntax clement according to given image sequences. Simulation results show that our proposed scheme outperforms the existing [7] me in compression efficiency without loss of quality.

Design of Hardwired Variable Length Decoder for H.264/AVC (하드웨어 구조의 H.264/AVC 가변길이 복호기 설계)

  • Yu, Yong-Hoon;Lee, Chan-Ho
    • Journal of the Institute of Electronics Engineers of Korea SD
    • /
    • v.45 no.11
    • /
    • pp.71-76
    • /
    • 2008
  • H.264(or MPEG-4/AVC pt.10) is a high performance video coding standard, and is widely used. Variable length code (VLC) of the H.264 standard compresses data using the statistical distribution of values. A decoder parses the compressed bit stream and searches decoded values in lookup tables, and the decoding process is not easy to implement by hardware. We propose an architecture of variable length decoder(VLD) for the H.264 baseline profile(BP) L4. The CAVLD decodes syntax elements using the combination of arithmetic units and lookup tables for the optimized hardware architecture. A barral shifter and a first 1's detector parse NAL bit stream, and are shared by Exp-Golomb decoder and CAVLD. A FIFO memory between CAVLD and the reorder unit and a buffer at the output of the reorder unit eliminate the bottleneck of data stream. The proposed VLD is designed using Verilog-HDL and is implemented using an FPGA. The synthesis result using a 0.18um standard CMOS technology shows that the gate count is 22,604 and the decoder can process HD($1920{\times}1080$) video at 120MHz.

A New Coding Method for Improving the Performance of MPEG-4 Part 10 Video Coding Standard (MPEG-4 Part 10 동영상 압축 표준 성능 개선을 위한 새로운 부호화 방식)

  • Moon, Yong-Ho
    • The Journal of Korean Institute of Communications and Information Sciences
    • /
    • v.31 no.11C
    • /
    • pp.1058-1065
    • /
    • 2006
  • In this paper, we propose a new motion vector coding algorithm suitable for MPEG-4 Part 10 video coding standard. In the proposed algorithm, the amount of the motion for a given video sequence is Predicted by using a characteristic of the motion vector distribution for the neighboring blocks as well as the MB_type, which is the syntax element in the standard. And one of the independent coding and the combined coding methods is adaptively employed to compress the motion vector difference. Simulation results show that the proposed algorithm outperforms the conventional methods without additional memory and calculations.