首页> 外文会议>International Conference on Advances in Electronics, Computers and Communications >VLSI implementation of discrete cosine transform and Intra prediction
【24h】

VLSI implementation of discrete cosine transform and Intra prediction

机译:离散余弦变换和帧内预测的VLSI实现

获取原文
获取外文期刊封面目录资料

摘要

This paper presents VLSI implementation of DCT algorithms for HEVC application. Here transform units (TU) of sizes 4×4 to 16×16 are implemented in both conventional method and partial butterfly model using multiplier-less multiple constant methods to reduce the hardware cost. Both the methods are designed using Verilog HDL, verified and compared for the realisation of hardware cost. We found that power and hardware utilization may be reduced by using MCM technique than conventional method. Along with DCT, Intra prediction algorithm is implemented. In the previous standards H.264, only 9 Intra prediction modes were used for the block size of 4×4 to 16×16. In the HEVC application the block sizes are increased from 4×4 to 64×64 and even the modes are increased to 34, to provide high performance gain. This is implemented using Verilog HDL. The functional verification, comparison and simulation are done using ISim simulator provided by Xilinx software. The synthesis results are obtained using RTL compiler provided by cadence software.
机译:本文介绍了用于HEVC应用的DCT算法的VLSI实现。此处,尺寸为4×4到16×16的变换单元(TU)在常规方法和部分蝶形模型中均使用无乘法器的多常数方法来实现,以降低硬件成本。两种方法均使用Verilog HDL设计,经过验证和比较以实现硬件成本。我们发现,与传统方法相比,使用MCM技术可能会降低功耗和硬件利用率。与DCT一起,实现了帧内预测算法。在以前的标准H.264中,对于4×4到16×16的块大小,仅使用了9种帧内预测模式。在HEVC应用中,块大小从4×4增加到64×64,甚至模式也增加到34,以提供高性能增益。这是使用Verilog HDL实现的。使用Xilinx软件提供的ISim仿真器进行功能验证,比较和仿真。使用cadence软件提供的RTL编译器可获得综合结果。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号