基于FPGA的2D-DCT处理器研究
|
摘要
离散余弦变换(DCT)在图像编解码方面应用十分广泛,至今已被JPEG、MPEG1、MPEG2、MPEG4和H26x等国际标准所采用。由于其计算量较大,软件实现往往难以满足实时处理的要求,因而在很多实际应用中需要采用硬件设计的DCT处理电路来满足我们对处理速度的要求。本文所研究的内容就是针对图像处理应用的8×8二维DCT处理核的硬件实现。
本文首先介绍了DCT在图像处理中的作用和原理,说明了利用DCT变换实现图像压缩的过程。接着,分析研究了DCT的各种快速算法,总结了前人对DCT快速算法及其实现所做的研究,并指出了其优点与不足。本文利用DCT的行列分离特性,将二维DCT实现转化为两个一维DCT并采用流水线设计技术实现。在一维DCT设计中,根据数据的特点对DA算法的数据进行了优化,通过利用数据的关联性,对占用的存储器数量进行压缩,并把乘法运算转化为移位求和,从而提高了流水线的执行速度,减少了资源占用量。最后,对所设计的DCT处理核进行了综合和时序仿真。
结果表明,当使用Altera公司MERCURY系列EP1M350F780C5型FPGA器件时,本文设计的方案能够在80M时钟频率下正确完成8×8的二维DCT的逻辑运算。
Discrete Cosine Transform are wildly used in coding and decoding of image processing. So far, it has adopted by many international standards like JPEG, MPEG1.MPEG2, MPEG4 and H26x etc. It's difficult to satisfy the requirement for real-time by software owing to its heavy quantity of computing. Therefore, we used to adopt hardware circuits to satisfy our requirement for speed in many practical applications. The content of the thesis researching is that hardware implementation of 8X8 2-D
DCT IP core for application of image processing.
The thesis introduced principle and effect of DCT in image processing at first, and elaborated the image processing which use DCT transform. And analyzed and studied various fast algorithms of DCT, summarized the working for DCT fast algorithms which has done by previous researchers. The thesis presented a way of DCT hardware implementation. The way adopted pipeline architecture and changed 2-D DCT to two 1-D DCT based on characteristic of row-column decomposition. In the design of 1-D DCT, we reduce the number of rom by optimizing the data and since we know the value of DCT cosine coefficients, by making use of its rotation characteristic, multiplication function can be designed by shift and addition logic (DA)instead of direct multiplication unit by which can save resources while improve the speed. At last, synthesize and verification for design has been done.
The result shown that the design could complete the function of 8X8 DCT under 80M clock frequency on the chip of EP1M350F780C5 which belongs to Altera company.
本文首先介绍了DCT在图像处理中的作用和原理,说明了利用DCT变换实现图像压缩的过程。接着,分析研究了DCT的各种快速算法,总结了前人对DCT快速算法及其实现所做的研究,并指出了其优点与不足。本文利用DCT的行列分离特性,将二维DCT实现转化为两个一维DCT并采用流水线设计技术实现。在一维DCT设计中,根据数据的特点对DA算法的数据进行了优化,通过利用数据的关联性,对占用的存储器数量进行压缩,并把乘法运算转化为移位求和,从而提高了流水线的执行速度,减少了资源占用量。最后,对所设计的DCT处理核进行了综合和时序仿真。
结果表明,当使用Altera公司MERCURY系列EP1M350F780C5型FPGA器件时,本文设计的方案能够在80M时钟频率下正确完成8×8的二维DCT的逻辑运算。
Discrete Cosine Transform are wildly used in coding and decoding of image processing. So far, it has adopted by many international standards like JPEG, MPEG1.MPEG2, MPEG4 and H26x etc. It's difficult to satisfy the requirement for real-time by software owing to its heavy quantity of computing. Therefore, we used to adopt hardware circuits to satisfy our requirement for speed in many practical applications. The content of the thesis researching is that hardware implementation of 8X8 2-D
DCT IP core for application of image processing.
The thesis introduced principle and effect of DCT in image processing at first, and elaborated the image processing which use DCT transform. And analyzed and studied various fast algorithms of DCT, summarized the working for DCT fast algorithms which has done by previous researchers. The thesis presented a way of DCT hardware implementation. The way adopted pipeline architecture and changed 2-D DCT to two 1-D DCT based on characteristic of row-column decomposition. In the design of 1-D DCT, we reduce the number of rom by optimizing the data and since we know the value of DCT cosine coefficients, by making use of its rotation characteristic, multiplication function can be designed by shift and addition logic (DA)instead of direct multiplication unit by which can save resources while improve the speed. At last, synthesize and verification for design has been done.
The result shown that the design could complete the function of 8X8 DCT under 80M clock frequency on the chip of EP1M350F780C5 which belongs to Altera company.
引文
[1]Jerry D.Gibson,Toby Berger,Tom Lookabaugh et al.多媒体数字压缩原理与标准.第一版.李煜晖,朱山风,段上为等译.北京:电子工业出版社,2000
[2]王云平,“DCT算法在图像处理中的作用,”辽宁工学院学报,2003:23(1),31-33
[3]Tekalp.A.M.著.崔之祜,江春,陈丽鑫译.数字视频处理.北京:电子工业出版社,1998
[4]沈兰荪,图像编码与异步传输.北京:人民邮电出版社,1998
[5]钟玉琢,乔秉新,祁卫译.运动图像及其伴音通用编码国际标准MPEG-2.北京:清华大学出版社,1997
[6]蔡士杰,岳华,刘小燕.连续色调静止图像的压缩与编码--JPEG.南京:南京大学出版社,1995
[7]N.Ahmed T.Natarajan andK.R.Rao "Discrete cosine transform "IEEE Trans.Comput,voi.C-23,Jan.1974,pp90-93
[8]W.H.Chen,C.H.Smith and S.C.Fralick,"A fast computational algorithm for the discrete cosine transform " IEEE Trans.Commun,vol.4,Sep.1977,pp1004-1009
[9]Zhongde Wang,"Fast Algorithms for The Discrete W Transform and for The Discrete fourier Transform," IEEE Trans.Signal Processing,vol.32,no.4,Aug.1984,pp803-816
[10]H.S.Hou "A fast recursive algorithm for computing the discrete cosine transform," IEEE Trans.Signal Processing,vol.35,no.10,Oct.1987,pp1455-1461
[11]马维祯,殷瑞祥,“DFT(2 m)和DCT(2 m)的快速递归新算法,”中国电子学会电路与系统年会论文集,深圳1987,26.20-26.23
[12]王中德,“计算离散cosine变换的一种算法,” 中国电子学会电路与系统年会论文集,深圳1987,26.9-26.12
[13]E.Feig and S.Winograd,"On theMultiplicative Complexity of Discrete Cosine Transforms," IEEE Trans.Inform.,vol.35,no.4,Apr.1992,pp1387-139l
[14]C.Loeffler,A.Ligtenberg,G.S.Moschytz,"Practical fast I-D DCT algorithms with ii multiplications," Proc.IEEE.ICASSP 89,1989,pp988-991
[15]郭斌.MPEG-2 压缩编码技术原理应用.http://info.broadcast.hc360.com/html/001/002/008/52120.htm 2004-4-1
[16]N.L.Cho and S.U.Lee,"Fast algorithm and implementation of 2-D discrete cosine transform," IEEE Trans.Circuits Syst.,vol.38,no.3,Mar.1991,pp297-305
[17]曾繁泰,李冰,李晓林.EDA工程概论.第一版.北京:清华大学出版社,2002
[18]A.S.Spanias,S.B.Jonsson,S.D.Stearns.Transform methods for seismic data compression.IEEE Transaction on Geoscience and Remote Sensing,1991,29(3),407-415
[19]Markus Puschel.Cooley-Tukey FFT like Algorithms for the DCT.Proc.ICASSP,2003,2:501-504
[20]A.Madisetti and A.N.Willson,A 100MHZ 2-D 8×8 DCT/IDCT processor for HDTV applications,IEEE Trans.Circuits Syst.Video Wechnol.,1995,5(2):158-164
[21]Docef.A,Kossentini.F.,Khanh Nquuyen-Phi et al,"The quantized DCT and its application to DCT-based video coding"IEEE Trans.Image Processing,vol.11,no.3,Mar.2002,pp177-187
[22]Byeong Lee.A New Algorithm to Compute the Discrete Cosine Transform.IEEE Transactions on Acoustics.Speech and Signal Processing,1984,32(6):1243-1245
[23]M.Vetterli,H.Nussbaumer.Simple FFT and DCT Algorithms with Reduced Number of Operations.Signal Processing(North Holland),1984,Vol 6:pp 267-278.
[24]N.Suehiro and M.Aatori.Fast Algorithms for the DFT and other Sinusoidal Transforms.IEEE Transactions on Acoustics,Speech,and Signal Processing,1986,34(3):642-644
[25]陈禾,毛志刚,叶以正.DCT快速算法及其VLSI实现.信号处理,1998,14:62-70
[26]P.Duhamel and H.H.Mida.New 2 DCT Algorithms suitable for VLSI Implementation.Proceedings IEEE International conference on Acoustics,Speech and Signal Processing,1987,12:1805-1808
[27]Liang Jie.Fast multiplierless approximations of the DCT with the lifting scheme.IEEE Transactions on Signal Processing,2001,49(12):3032-3044
[28]张海亮.浅议电视信号的数字化与码率压缩.http://www.lunw.com/thesis/39/7043_1.html,1999-2-1
[29]Hassan EL-Banna,Alaa A.EL-Fattah.An Efficient Implementation of the 1D-DCT using FPGA Technology.ECBS 04,2004,356-360
[30]Y Arai,T Agui,M Nakajima.A fast DCT-SQ scheme for images.The Transaction of The IEIGE,1988,71(11):1095-1097
[31]Ephraim Feig.A fast scaled DCT algorithm.Proceedings of SPIE,1990,1244:2-13
[32]Lizhi Cheng,Yonghong Zeng.New polynomial transform algorithm for multidimensional type-Ⅳ DCTs.IEEE Trans Signal Process,2000,48(10):2814-2821
[33]J.Prado,P.Duhamel.A polynomial transform based computation of the 2-D DCT with minimum multiplicative complexity.ICASSP,1996,Vol 3:pp 1347-1350
[34]P.Duhamel,C.Guillemot.Polynomial transform computation of the 2-D DCT.Acoustics,Speech,and Signal Processing,1990,Vol 3:pp 1515-1518
[35]http://ffmpeg.sourceforge.net/index.php,2005-10-1
[36]http://www.xvid.org/,2005-11-1
[37]魏忠义,朱磊.基于DSP的JPEG图像解码算法的实现.现代电子技术.2005,Vol2:pp 66-68
[38]刘宝兰,刘贵忠,苏睿.H.264中整数DCT变换及量化的DSP实现.微电子学与计算机,2005,Vol 6:pp 200-205
[39]金燕波,朗锐,罗发根等.利用TMS320C6201 DSP芯片进行图像压缩.电子技术应用,2004,Vol 1:pp 63-66
[40]陈玲晶,郑学仁,范健民等.二维DCT/IDCT的FPGA实现及验证方法.集成电路应用,2005,Vol 1:pp 49-53
[41]钟文荣,陈建发.二维DCT算法的高速芯片设计.厦门大学学报:自然科学版,2005,Vol 2:pp 198-201
[42]http://www.avs.org.cn/,2005-08-01
[43]陈禾,韩月秋,基于改进DA算法的2-D IDCT系统设计.通信学报,2001.3:80-85
[44]山洪刚等.一种应用于8×8二维DCT/IDCT的高效结构.半导体技术,2002.6:13-17
[45]杨依忠.IP核及其集成技术的研究.硕士论文,合肥工业大学,2004.4
[46]张光烈.实时视频处理算法及其IP核的VLSI实现.博士论文,西安交通大学,2002.10
[47]W.-J.Duh and J.-L.Wu,Constant-rotation DCT architecture based on CORDIC techniques,INT.J.ELECTRONIC,1990,69(5):583-593
[48]E.P.Mariatos,D.E.Metafas,J.A.Hallas and C.E.Goutis,A fast DCT processor,based on special purpose CORDIC rotators,1994 IEEE International Symposium on Circuit and Systems.,1994,vol(6):271-274
[49]F.Zhou and P.Kornerup,High speed DCT/IDCT using a pipeline CORDIC algorithm,(i063-6889/95 1995 IEEE),1995,vol(7):180-187
[50]C.-T.Chiu and K.J.R.Liu,Real-time parallel and fully pipeline two-dimensional DCT lattice structures with application to HDTV systems,IEEE Trans.Circuit Syst.Video Technol.,1992,2(1):25-36
[51]V.Srinivasan and K.J.R.hiu,VLSI design of high-speed time-recursive 2-D DCT/IDCT processor for video applications,IEEE Trans.Circuit syst.Video Technol.,1996,6(1)
[52]E.Feig,A fast scaled DCT algorithm,Proc.SPEC Int.$oc.Opt.Eng.,1990,vol.1244,2-13
[2]王云平,“DCT算法在图像处理中的作用,”辽宁工学院学报,2003:23(1),31-33
[3]Tekalp.A.M.著.崔之祜,江春,陈丽鑫译.数字视频处理.北京:电子工业出版社,1998
[4]沈兰荪,图像编码与异步传输.北京:人民邮电出版社,1998
[5]钟玉琢,乔秉新,祁卫译.运动图像及其伴音通用编码国际标准MPEG-2.北京:清华大学出版社,1997
[6]蔡士杰,岳华,刘小燕.连续色调静止图像的压缩与编码--JPEG.南京:南京大学出版社,1995
[7]N.Ahmed T.Natarajan andK.R.Rao "Discrete cosine transform "IEEE Trans.Comput,voi.C-23,Jan.1974,pp90-93
[8]W.H.Chen,C.H.Smith and S.C.Fralick,"A fast computational algorithm for the discrete cosine transform " IEEE Trans.Commun,vol.4,Sep.1977,pp1004-1009
[9]Zhongde Wang,"Fast Algorithms for The Discrete W Transform and for The Discrete fourier Transform," IEEE Trans.Signal Processing,vol.32,no.4,Aug.1984,pp803-816
[10]H.S.Hou "A fast recursive algorithm for computing the discrete cosine transform," IEEE Trans.Signal Processing,vol.35,no.10,Oct.1987,pp1455-1461
[11]马维祯,殷瑞祥,“DFT(2 m)和DCT(2 m)的快速递归新算法,”中国电子学会电路与系统年会论文集,深圳1987,26.20-26.23
[12]王中德,“计算离散cosine变换的一种算法,” 中国电子学会电路与系统年会论文集,深圳1987,26.9-26.12
[13]E.Feig and S.Winograd,"On theMultiplicative Complexity of Discrete Cosine Transforms," IEEE Trans.Inform.,vol.35,no.4,Apr.1992,pp1387-139l
[14]C.Loeffler,A.Ligtenberg,G.S.Moschytz,"Practical fast I-D DCT algorithms with ii multiplications," Proc.IEEE.ICASSP 89,1989,pp988-991
[15]郭斌.MPEG-2 压缩编码技术原理应用.http://info.broadcast.hc360.com/html/001/002/008/52120.htm 2004-4-1
[16]N.L.Cho and S.U.Lee,"Fast algorithm and implementation of 2-D discrete cosine transform," IEEE Trans.Circuits Syst.,vol.38,no.3,Mar.1991,pp297-305
[17]曾繁泰,李冰,李晓林.EDA工程概论.第一版.北京:清华大学出版社,2002
[18]A.S.Spanias,S.B.Jonsson,S.D.Stearns.Transform methods for seismic data compression.IEEE Transaction on Geoscience and Remote Sensing,1991,29(3),407-415
[19]Markus Puschel.Cooley-Tukey FFT like Algorithms for the DCT.Proc.ICASSP,2003,2:501-504
[20]A.Madisetti and A.N.Willson,A 100MHZ 2-D 8×8 DCT/IDCT processor for HDTV applications,IEEE Trans.Circuits Syst.Video Wechnol.,1995,5(2):158-164
[21]Docef.A,Kossentini.F.,Khanh Nquuyen-Phi et al,"The quantized DCT and its application to DCT-based video coding"IEEE Trans.Image Processing,vol.11,no.3,Mar.2002,pp177-187
[22]Byeong Lee.A New Algorithm to Compute the Discrete Cosine Transform.IEEE Transactions on Acoustics.Speech and Signal Processing,1984,32(6):1243-1245
[23]M.Vetterli,H.Nussbaumer.Simple FFT and DCT Algorithms with Reduced Number of Operations.Signal Processing(North Holland),1984,Vol 6:pp 267-278.
[24]N.Suehiro and M.Aatori.Fast Algorithms for the DFT and other Sinusoidal Transforms.IEEE Transactions on Acoustics,Speech,and Signal Processing,1986,34(3):642-644
[25]陈禾,毛志刚,叶以正.DCT快速算法及其VLSI实现.信号处理,1998,14:62-70
[26]P.Duhamel and H.H.Mida.New 2 DCT Algorithms suitable for VLSI Implementation.Proceedings IEEE International conference on Acoustics,Speech and Signal Processing,1987,12:1805-1808
[27]Liang Jie.Fast multiplierless approximations of the DCT with the lifting scheme.IEEE Transactions on Signal Processing,2001,49(12):3032-3044
[28]张海亮.浅议电视信号的数字化与码率压缩.http://www.lunw.com/thesis/39/7043_1.html,1999-2-1
[29]Hassan EL-Banna,Alaa A.EL-Fattah.An Efficient Implementation of the 1D-DCT using FPGA Technology.ECBS 04,2004,356-360
[30]Y Arai,T Agui,M Nakajima.A fast DCT-SQ scheme for images.The Transaction of The IEIGE,1988,71(11):1095-1097
[31]Ephraim Feig.A fast scaled DCT algorithm.Proceedings of SPIE,1990,1244:2-13
[32]Lizhi Cheng,Yonghong Zeng.New polynomial transform algorithm for multidimensional type-Ⅳ DCTs.IEEE Trans Signal Process,2000,48(10):2814-2821
[33]J.Prado,P.Duhamel.A polynomial transform based computation of the 2-D DCT with minimum multiplicative complexity.ICASSP,1996,Vol 3:pp 1347-1350
[34]P.Duhamel,C.Guillemot.Polynomial transform computation of the 2-D DCT.Acoustics,Speech,and Signal Processing,1990,Vol 3:pp 1515-1518
[35]http://ffmpeg.sourceforge.net/index.php,2005-10-1
[36]http://www.xvid.org/,2005-11-1
[37]魏忠义,朱磊.基于DSP的JPEG图像解码算法的实现.现代电子技术.2005,Vol2:pp 66-68
[38]刘宝兰,刘贵忠,苏睿.H.264中整数DCT变换及量化的DSP实现.微电子学与计算机,2005,Vol 6:pp 200-205
[39]金燕波,朗锐,罗发根等.利用TMS320C6201 DSP芯片进行图像压缩.电子技术应用,2004,Vol 1:pp 63-66
[40]陈玲晶,郑学仁,范健民等.二维DCT/IDCT的FPGA实现及验证方法.集成电路应用,2005,Vol 1:pp 49-53
[41]钟文荣,陈建发.二维DCT算法的高速芯片设计.厦门大学学报:自然科学版,2005,Vol 2:pp 198-201
[42]http://www.avs.org.cn/,2005-08-01
[43]陈禾,韩月秋,基于改进DA算法的2-D IDCT系统设计.通信学报,2001.3:80-85
[44]山洪刚等.一种应用于8×8二维DCT/IDCT的高效结构.半导体技术,2002.6:13-17
[45]杨依忠.IP核及其集成技术的研究.硕士论文,合肥工业大学,2004.4
[46]张光烈.实时视频处理算法及其IP核的VLSI实现.博士论文,西安交通大学,2002.10
[47]W.-J.Duh and J.-L.Wu,Constant-rotation DCT architecture based on CORDIC techniques,INT.J.ELECTRONIC,1990,69(5):583-593
[48]E.P.Mariatos,D.E.Metafas,J.A.Hallas and C.E.Goutis,A fast DCT processor,based on special purpose CORDIC rotators,1994 IEEE International Symposium on Circuit and Systems.,1994,vol(6):271-274
[49]F.Zhou and P.Kornerup,High speed DCT/IDCT using a pipeline CORDIC algorithm,(i063-6889/95 1995 IEEE),1995,vol(7):180-187
[50]C.-T.Chiu and K.J.R.Liu,Real-time parallel and fully pipeline two-dimensional DCT lattice structures with application to HDTV systems,IEEE Trans.Circuit Syst.Video Technol.,1992,2(1):25-36
[51]V.Srinivasan and K.J.R.hiu,VLSI design of high-speed time-recursive 2-D DCT/IDCT processor for video applications,IEEE Trans.Circuit syst.Video Technol.,1996,6(1)
[52]E.Feig,A fast scaled DCT algorithm,Proc.SPEC Int.$oc.Opt.Eng.,1990,vol.1244,2-13