计算机科学与探索 ›› 2008, Vol. 2 ›› Issue (3): 311-320.

• 学术研究 • 上一篇    下一篇

数据驱动的并行视频编码器的设计和实现

于晓艳+,荣宪伟   

  1. 哈尔滨师范大学 物理系,哈尔滨 150080
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2008-06-20 发布日期:2008-06-20
  • 通讯作者: 于晓艳

Design and implementation of a data-driven parallel video codec

YU Xiaoyan+, RONG Xianwei   

  1. Department of Physics, Harbin Normal University, Harbin 150080, China
  • Received:1900-01-01 Revised:1900-01-01 Online:2008-06-20 Published:2008-06-20
  • Contact: YU Xiaoyan

摘要: 通用视频编码系统一般基于标准视频压缩算法,比如MPEG4和H.264。但这些标准算法包括如离散余弦变换和运动估计等高复杂度的计算,所以为了实现实时通信,编码系统通常采用以硬件为基础的实现方法。然而以硬件为基础的视频编码系统具有实施周期长、成本高、灵活性差等不足。文章介绍了一种基于低计算复杂度自适应块截短编码方案的数据驱动并行视频编码系统,实验结果验证了编码系统的有效性。

关键词: 数据驱动的实施, 自适应块截短编码, 视频编码器

Abstract: Generic video codec systems are based upon standard compression algorithms such as MPEG4 and H.264. Since these standard algorithms include high-complexity computation like discrete cosine transform and motion estimation, they are generally implemented by hardware-based methods for real-time applications. However, such kinds of codec systems endure long-period implementation, expensive spending and poor flexibility because they are hardwired. This paper introduces a data-driven parallel video codec system that uses a low-complexity Adaptive Block Truncation-Coding (ABTC) scheme. The efficiency of the codec system introduced is confirmed using standard video sequences.

Key words: data-driven implementation, adaptive block truncation coding, video codec