期刊文献+

二维DCT快速算法及FPGA实现 被引量:10

Fast 2-D Discrete Cosine Transform Algorithm and Its Implementation Based On FPGA
下载PDF
导出
摘要 本文提出了一种二维DCT快速算法的FPGA实现结构,采用行列快速算法将二维DCT分解成两个一维DCT实现,其中一维DCT借鉴Loeffler DCT算法,采用并行的流水线结构,提高电路的数据吞吐率和运算速度,通过系数矩阵的简化和蝶形运算结构的等价减少乘法器的消耗,一维DCT核消耗16个乘法器。转置RAM采用8片双口RAM,一个时钟可以完成8个数据读写。实验结果验证了二维DCT核设计的正确性,该电路结构消耗资源少,布线简单,功耗小,适合图像的实时处理。 In this paper a fast Discrete Cosine Transform (DCT) implementation architecture using FPGA is presented. The architecture uses a row-column decomposition to implement a 2-D DCT using twO 1-D DCT in series, The 1-D DCT processor is realized by optimizing the Loeffler DCT with parallel- pipeline architecture, which makes the circuit work with high data throughput and speed, and it reduces the number of multipliers by .simplifying the coefficient matrix and modifying the butterfly architecture, The 1-D DCT processor requires 16 multipliers, The transposition memory uses 8 Dual-Port rams, which can do eight reads and writes in one clock, Experimental results show that the design is correct. The circuit structure has the advantages of small chip size, simple wiring and low-power, which is extremely suitable to real-time image processing
出处 《电子质量》 2008年第2期5-7,22,共4页 Electronics Quality
关键词 二维DCT算法 流水线 转置存储器 FPGA 2-D DCT Pipeline Transposition Memory FPGA
  • 相关文献

参考文献1

二级参考文献5

  • 1Domagoj Babic. Discrete Cosine Transform Algorithms for FPGA Devices[M]. Zagreb, 2003.
  • 2Chris H Dick. Minimum Multiplicative Complexity Implementation of the 2 - D DCT Using Xilinx FPGAs[J]. Proc.of SPIE' s Photonies East'98 Configurable Computing, Technology and Applications, Boston, MA, USA, 1998, 3526(23) : 190 - 201.
  • 3Xilinx. Virtex - Ⅱ Platform FPGAs Complete Data Sheet[EB]. http:// direct. xilinx.com/ hvdoes/puhlications/ds031.pdf,2005 - 03 - 01/2005 - 11 - 02.
  • 4山洪刚,郑南宁,杨国安,张光烈.一种应用于8×8二维DCT/IDCT的高效结构[J].半导体技术,2002,27(6):13-17. 被引量:8
  • 5崔春艳,李彩霞.基于DCT变换的数字图像压缩技术及其Matlab实现[J].现代电子技术,2002,25(9):7-9. 被引量:7

共引文献2

同被引文献58

引证文献10

二级引证文献31

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部