卫星图像实时压缩设备中的关键技术研究
|
摘要
随着数字化技术的发展和信息化进程的不断深入，图像压缩技术成为一个独立的、重要的研究领域。卫星图像压缩技术不同于普通的静态图像压缩，它在硬件上所受的约束更多，而实时性要求更高。本文主要设计并实现了一种基于双正交整数叠式变换(LBT)和简化零树编码的实时卫星图像压缩方法，进行了相应的实验研究。
目前图像压缩中研究和应用最为广泛的是基于离散小波变换(DWT)的图像压缩方法。良好的能量集中特性、多分辨率特性以及变换系数中与零树编码紧密结合的零树结构特性，使小波变换在静态图像压缩中表现出优异的性能，编码质量比以往的方法明显提高，并且自然地实现了码流的渐进传输，能够满足不同类型的需求。然而过大的计算量和过高的存储需求使小波变换不适于应用在卫星图像压缩设备以及一些小型便携设备中。为此，采用一种基于LBT的低复杂、低存储需求的图像压缩方法。在采用二进算法时，LBT所需的计算量仅为CDF9／7小波的38％，所需存储开销更是不到CDF9／7小波的1／15。通过分析LBT变换系数的统计分布特点，采用一种简化的零树编码方法对变换系数进行编码，使算法能够适应星载图像压缩。
在算法研究的基础上，本文探讨了该图像压缩算法的硬件实现问题，包括系统的硬件设计和软件设计。硬件设计的总体思想是采用多DSP+FPGA的设计结构，解决了传统多FPGA设计成本过高以及单粒子逻辑门翻转效应等问题。软件设计着重考虑了实时压缩软件的代码优化问题。
通过标准卫星遥感样本图像的压缩测试实验，结果表明，我们研究并实现的基于LBT的图像压缩算法在恢复图像质量等综合性能方面优于基于JPEG2000的图像压缩算法，且具有计算简便、存储要求低等特点，在实时性上体现出不可比拟的优势，在卫星图像数据压缩领域具有重要的应用价值。
As the development of digital technology, image compression becomes a separateimportant course to be researched. Satellite image compression technology differs fromnormal image compression technology, because it is more restricted on hardwarecondition,and requires further real-time characteristic. This paper mainly discusses theimage compression system based on lapped biothorgonal transform (LBT) andsimplified zero tree codec, and its realization.
As we know, today the discrete wavelet transform (DWT) based imagecompression method is most widely used. The well ability of energy centralizing, themultiresolution characteristic and the zero tree construction of the transformed modulusmake the DWT based method perform well in the area of static image compression. Thecoding efficiency is higher than any other former method, and it is naturally gradationalto match multi-level requirement. However, DWT is not fit for satellite imagecompression and some portable equipment because of its high computation and highmemory requirement. Because of this, the paper brings forward the LBT based lowcomputation and low memory needed image compression method. In the method, LBTcosts only 38％computation as DWT, and the memory cost is even lower, only 1/15 ofDWT. Through the research of the characteristics of the LBT transform modulus, a typeof simplified zero tree codec is used the method, making the arithmetic fitable forsatellite image compression.
Based on the LBT method, the paper discusses the realization of the satellite imagecompression experimental system, including the hardware design as well as the softwaredesign. The hardware is mainly constructed of multiple DSP and FPGA, which solvesthe high cost and low reliability of former designs based on multi FPGA. Theoptimization of the effecience of the code is emphasized in the software design.
The experiment of standard test image compression indicates that, the performanceof the LBT based image compression system we designed, is already better than theJPEG2000 method in the aspect of the quality of resumed image. As our method ismuch easier to realize because of its low computation and low memory requirementcharacteristics, it bares much more practicability in the area of satellite imagecompression.
目前图像压缩中研究和应用最为广泛的是基于离散小波变换(DWT)的图像压缩方法。良好的能量集中特性、多分辨率特性以及变换系数中与零树编码紧密结合的零树结构特性，使小波变换在静态图像压缩中表现出优异的性能，编码质量比以往的方法明显提高，并且自然地实现了码流的渐进传输，能够满足不同类型的需求。然而过大的计算量和过高的存储需求使小波变换不适于应用在卫星图像压缩设备以及一些小型便携设备中。为此，采用一种基于LBT的低复杂、低存储需求的图像压缩方法。在采用二进算法时，LBT所需的计算量仅为CDF9／7小波的38％，所需存储开销更是不到CDF9／7小波的1／15。通过分析LBT变换系数的统计分布特点，采用一种简化的零树编码方法对变换系数进行编码，使算法能够适应星载图像压缩。
在算法研究的基础上，本文探讨了该图像压缩算法的硬件实现问题，包括系统的硬件设计和软件设计。硬件设计的总体思想是采用多DSP+FPGA的设计结构，解决了传统多FPGA设计成本过高以及单粒子逻辑门翻转效应等问题。软件设计着重考虑了实时压缩软件的代码优化问题。
通过标准卫星遥感样本图像的压缩测试实验，结果表明，我们研究并实现的基于LBT的图像压缩算法在恢复图像质量等综合性能方面优于基于JPEG2000的图像压缩算法，且具有计算简便、存储要求低等特点，在实时性上体现出不可比拟的优势，在卫星图像数据压缩领域具有重要的应用价值。
As the development of digital technology, image compression becomes a separateimportant course to be researched. Satellite image compression technology differs fromnormal image compression technology, because it is more restricted on hardwarecondition,and requires further real-time characteristic. This paper mainly discusses theimage compression system based on lapped biothorgonal transform (LBT) andsimplified zero tree codec, and its realization.
As we know, today the discrete wavelet transform (DWT) based imagecompression method is most widely used. The well ability of energy centralizing, themultiresolution characteristic and the zero tree construction of the transformed modulusmake the DWT based method perform well in the area of static image compression. Thecoding efficiency is higher than any other former method, and it is naturally gradationalto match multi-level requirement. However, DWT is not fit for satellite imagecompression and some portable equipment because of its high computation and highmemory requirement. Because of this, the paper brings forward the LBT based lowcomputation and low memory needed image compression method. In the method, LBTcosts only 38％computation as DWT, and the memory cost is even lower, only 1/15 ofDWT. Through the research of the characteristics of the LBT transform modulus, a typeof simplified zero tree codec is used the method, making the arithmetic fitable forsatellite image compression.
Based on the LBT method, the paper discusses the realization of the satellite imagecompression experimental system, including the hardware design as well as the softwaredesign. The hardware is mainly constructed of multiple DSP and FPGA, which solvesthe high cost and low reliability of former designs based on multi FPGA. Theoptimization of the effecience of the code is emphasized in the software design.
The experiment of standard test image compression indicates that, the performanceof the LBT based image compression system we designed, is already better than theJPEG2000 method in the aspect of the quality of resumed image. As our method ismuch easier to realize because of its low computation and low memory requirementcharacteristics, it bares much more practicability in the area of satellite imagecompression.
引文
[1] N. Ahmed, T. Natarajan, K. Rao. Discrete cosine transform. IEEE Transactions Computer. Jan. 1974, 23: 88-93
[2] H. S. Malvar, D. H. Staelin. The LOT: Transform coding without blocking effects. IEEE Transactions on Acoust Speech, signal processing. 1987, 37(4): 553-559
[3] H. S. Malvar. Biorthogonal and nonuniform lapped transforms for transform coding with reduce blocking and ringing artifacts. IEEE Transactions on Signal Processing, 1998, 46(4): 1043-1053
[4] C. Tu, T. D. Tran. Context-based entropy coding of lock transform coefficients for image compression. IEEE Transactions on Image Processing. 2002, 11(11): 1271-1283
[5] Jie L., T. D. Tran. Fast multiplierless approximation of the DCT with lift scheme. IEEE Transactions on Signal Processing. 2001, 49(2): 3032-3044
[6] W. Sweldens. The lifting scheme: A new philosophy in biorthogonal wavelet constructions. Proc SPIE. 1995, 2569: 68-79
[7] Michael David Adams. Reversible integer-to integer wavelet transforms for image coding. Ph. D. thesis, The university of British Columbia, Vancouver, Canada, September 2002
[8] 钟广军，成礼智，陈火旺．双正交重叠变换的整数实现算法与图像压缩．电子学报．Nov．2001，29(11)：2001
[9] J. Shapiro. Embedded image coding using zerotrees of wavelet coefficients. IEEE Transactions on Signal Processing. 1993, 41 (12): 3445-3462
[10] A. Said, W. A. Pearlman. A new, fast, and efficient image codec based on set partitioning in hierarchical trees. IEEE Transactions on Circuit and System for Video Technology. June 1996, 6(3): 243-250
[11] D. Taubman. High performance scalable image compression with EBCOT. IEEE Transactions on Image Processing. July 2000, 9(7): 1158-1170
[12] ISO/IEC/JTC1/SC20 WG1 N2000. JPEG2000 partl final committee draft. December 2000.
[13] David S. Taubman, Michael W. Marcellin. JPEG2000 image compression fundamentals, standards, and practice. THE KLUWER INTERNATIONAL SERIES IN ENGINEERING AND COMPUTER SCIENCE, Boston/Dordrecht/London: Kluwer Academic Publisher, 2002
[14] C. Christopoulos, J. Askelof, M. Larsson. Efficient methods for encoding regions of interest in the upcoming JPEG2000 still image coding standard. IEEE Signal Processing Letters. September 2000, 7(9): 247-249
[15] R. Grosbois, D. S.-C., T. Ebrahimi. New approach to JPEG2000 compliant region of interest coding. Proc. of the SPIE's 46th Annual Meeting, Applications of Digital Imaging Processing ⅩⅩⅣ. San Diego, CA, USA, 2001, vol. 4172
[16] W. A. Pearlman, A. Islam, N. Nagaraj, A. Said. Efficient, low-complexity image coding with a set-partitioning embedded block coder, 2003
[17] B. Brian, T. R. Fischer. Quadtree classification and TCQ image coding. J. A. Storer and M. Cohn, (Editor) DCC'99: Data Compression Conference. Los Alamitos, CA: IEEE Computer Society, 1999, 149-157
[18] A. Zandi, J. Allen, E. Schwartz, M. Boliek. CREW: Compression with recersible embedded wavelets. J. A. Storer adn M. Cohn, (Editor) DCC'99: Data Compression Conference. Los Alamitos, CA: IEEE Computer Society, 1999, 212-221
[19] C. Chrysafis, A. Ortega. Line based, reduced memory, wavelet image compression. IEEE Transactions on Image Processing. March 2000, 9(3): 378-389
[20] Haffner Patrick et. High-quality document image compression with DjVu. Journal of Electronic Imaging, SPIE. 1998, 7(3): 410-425
[21] Brislawn Christopher, Jonathan Bradley, R. Onyschczak, Tom Hopper. The FBI ompression standard for digitized fingerprint images. Proceedings SPTE. Auguest 1996, 2847: 344-355
[22] I. Daubechies. Ten lectures on wavelets. SIAM, Philadelphia, PA, 1992
[23] I. Daubechies, W. Sweldens. Factoring wavelet transforms into lifting steps. J Fourier Anal Appl. 1998, 4(3): 245-267
[24] Liang Diannong, Zhang Zhenghui Cheng Lizhi. The general construction of the biorthogonal wavelet and applications in image coding. Opt Eng. 2003, 2(7): 1949-1955
[25] T. D. Tran, T. Q. Nguyen. A progressive transmission image coder using linear phase uniform filterbanks as block transform. IEEE Transactions on Image Processing. 1999, 8(11): 1493-1507
[26] R. L. de Queiroz, T. Q. Nguyen, K. R. Rao. The GenLOT: Generalized linear-phase lapped orthogonal transform. IEEE Transactions on Signal Processing. 1996: 497-507
[27] 成礼智．离散与小波变换新型算法及其在图像处理中应用的研究．Ph．D．thesis，国防科学技术大学，长沙．Dec．2002
[28] Malvar H. S. Signal processing with lapped transforms. NJ: Prentice-Hall, 1992
[29] 陈波．基于变换的低复杂度低存储需求的图像压缩方法研究．国防科技大学硕士学位论文．Nov．2004
[30] T. D. Tran, Jie L., C. Tu. Lapped transform via time-domain pre- and post-filtering. IEEE Transactions on Signal Processing. 2003
[31] B. G. Lee. A new algorithm to compute the discrete cosine transform. IEEE Transactions ASSP-32. 1984, (6)
[32] T. D. Tran. The binDCT: fast multiplierless approximation of the DCT. IEEE Signal Processing Letters. Jun 2000, 7(6): 141-144
[33] F. W. Wheeler, W. A. Pearlman. SPIHT image compression without lists. Proc. of the International Conf. on Acoustics, Speech and Signal Processing. Vancouver, Canada, 2000, 2047-2050
[34] F. W. Wheeler, W. A. Pearlman. Low memory packetized spiht image compression. M. B. Mattthews, (Editor) Conf. Record of The Thirty-Third Asilomar Conf. on Signals, Systems and Computers(M. B. Matthews, ed.). Pacific Grove, CA, 1999, vol. 2, 1193-1197
[35] C. Y. Su, B. F. Wu. A low memory zerotree coding for arbitrarily shaped objects. IEEETransactions on Image Processing. March 2003, 12(3): 271-282
[36] Yong Sun, Hui Zhang, Guangshu Hu. Real-time implementation of a new low-memory SPIHT image coding algorithm using dsp chip. IEEE Transactions on Image Processing. September 2002, 11 (9): 1112-1115
[37] J. M. Shapiro. A fast technique for identifying zerotrees in the EZW algorithm. Proc. of the International Conf. on Acoustics, Speech and Signal Processing. 1996, 1455-1458
[38] Zhao Debin, Gao Wen. Low complexity and low memory entropy coder for image compression. IEEE Transactions on Circuit and System for Video Technology. 2001, 11 (10):1140-1145
[39] Aaron Deever, Sheila Hemami. Efficient sign coding and estimation of zero-quantized coefficients in embedded wavelet image codecs. IEEE Transactions Image Processing. 2003, 12(4):420-429
[40] A. Said, W. A. Pearlman. SPIHT image compression home page. Online, 1996. www.cipr.rpi.edu/research/SPIHT/
[41] 成礼智，王洪霞，罗永．小波的理论与应用．研究生教材丛书，科学出版社，2003
[42] 曾文曲，文有为，孙炜．分形小波与图像压缩．东北大学出版社，2002
[43] 周娟，罗建书，乔士东．结合小波变换的零搜索分型图像编码．中国图像图形学报．July 2001，6(7)
[44] 冯晋臣，赵荣堂．陆地卫星CCT数据压缩处理研究．南京林产工业学院学报．1981．3：98-114
[45] 张杨．空间太阳望远镜中图像压缩单元软件设计．山东科技大学学报．2003，22(1)
[46] 钟是等．小波变换在遥感图像压缩中应用研究．中国图形图像学报．October 1998，3(10)
[47] 郭武，梅丽，罗建书．一种基于SPIHT的ROI图形编码．中国空间科学技术．Feb 2003(1)
[2] H. S. Malvar, D. H. Staelin. The LOT: Transform coding without blocking effects. IEEE Transactions on Acoust Speech, signal processing. 1987, 37(4): 553-559
[3] H. S. Malvar. Biorthogonal and nonuniform lapped transforms for transform coding with reduce blocking and ringing artifacts. IEEE Transactions on Signal Processing, 1998, 46(4): 1043-1053
[4] C. Tu, T. D. Tran. Context-based entropy coding of lock transform coefficients for image compression. IEEE Transactions on Image Processing. 2002, 11(11): 1271-1283
[5] Jie L., T. D. Tran. Fast multiplierless approximation of the DCT with lift scheme. IEEE Transactions on Signal Processing. 2001, 49(2): 3032-3044
[6] W. Sweldens. The lifting scheme: A new philosophy in biorthogonal wavelet constructions. Proc SPIE. 1995, 2569: 68-79
[7] Michael David Adams. Reversible integer-to integer wavelet transforms for image coding. Ph. D. thesis, The university of British Columbia, Vancouver, Canada, September 2002
[8] 钟广军，成礼智，陈火旺．双正交重叠变换的整数实现算法与图像压缩．电子学报．Nov．2001，29(11)：2001
[9] J. Shapiro. Embedded image coding using zerotrees of wavelet coefficients. IEEE Transactions on Signal Processing. 1993, 41 (12): 3445-3462
[10] A. Said, W. A. Pearlman. A new, fast, and efficient image codec based on set partitioning in hierarchical trees. IEEE Transactions on Circuit and System for Video Technology. June 1996, 6(3): 243-250
[11] D. Taubman. High performance scalable image compression with EBCOT. IEEE Transactions on Image Processing. July 2000, 9(7): 1158-1170
[12] ISO/IEC/JTC1/SC20 WG1 N2000. JPEG2000 partl final committee draft. December 2000.
[13] David S. Taubman, Michael W. Marcellin. JPEG2000 image compression fundamentals, standards, and practice. THE KLUWER INTERNATIONAL SERIES IN ENGINEERING AND COMPUTER SCIENCE, Boston/Dordrecht/London: Kluwer Academic Publisher, 2002
[14] C. Christopoulos, J. Askelof, M. Larsson. Efficient methods for encoding regions of interest in the upcoming JPEG2000 still image coding standard. IEEE Signal Processing Letters. September 2000, 7(9): 247-249
[15] R. Grosbois, D. S.-C., T. Ebrahimi. New approach to JPEG2000 compliant region of interest coding. Proc. of the SPIE's 46th Annual Meeting, Applications of Digital Imaging Processing ⅩⅩⅣ. San Diego, CA, USA, 2001, vol. 4172
[16] W. A. Pearlman, A. Islam, N. Nagaraj, A. Said. Efficient, low-complexity image coding with a set-partitioning embedded block coder, 2003
[17] B. Brian, T. R. Fischer. Quadtree classification and TCQ image coding. J. A. Storer and M. Cohn, (Editor) DCC'99: Data Compression Conference. Los Alamitos, CA: IEEE Computer Society, 1999, 149-157
[18] A. Zandi, J. Allen, E. Schwartz, M. Boliek. CREW: Compression with recersible embedded wavelets. J. A. Storer adn M. Cohn, (Editor) DCC'99: Data Compression Conference. Los Alamitos, CA: IEEE Computer Society, 1999, 212-221
[19] C. Chrysafis, A. Ortega. Line based, reduced memory, wavelet image compression. IEEE Transactions on Image Processing. March 2000, 9(3): 378-389
[20] Haffner Patrick et. High-quality document image compression with DjVu. Journal of Electronic Imaging, SPIE. 1998, 7(3): 410-425
[21] Brislawn Christopher, Jonathan Bradley, R. Onyschczak, Tom Hopper. The FBI ompression standard for digitized fingerprint images. Proceedings SPTE. Auguest 1996, 2847: 344-355
[22] I. Daubechies. Ten lectures on wavelets. SIAM, Philadelphia, PA, 1992
[23] I. Daubechies, W. Sweldens. Factoring wavelet transforms into lifting steps. J Fourier Anal Appl. 1998, 4(3): 245-267
[24] Liang Diannong, Zhang Zhenghui Cheng Lizhi. The general construction of the biorthogonal wavelet and applications in image coding. Opt Eng. 2003, 2(7): 1949-1955
[25] T. D. Tran, T. Q. Nguyen. A progressive transmission image coder using linear phase uniform filterbanks as block transform. IEEE Transactions on Image Processing. 1999, 8(11): 1493-1507
[26] R. L. de Queiroz, T. Q. Nguyen, K. R. Rao. The GenLOT: Generalized linear-phase lapped orthogonal transform. IEEE Transactions on Signal Processing. 1996: 497-507
[27] 成礼智．离散与小波变换新型算法及其在图像处理中应用的研究．Ph．D．thesis，国防科学技术大学，长沙．Dec．2002
[28] Malvar H. S. Signal processing with lapped transforms. NJ: Prentice-Hall, 1992
[29] 陈波．基于变换的低复杂度低存储需求的图像压缩方法研究．国防科技大学硕士学位论文．Nov．2004
[30] T. D. Tran, Jie L., C. Tu. Lapped transform via time-domain pre- and post-filtering. IEEE Transactions on Signal Processing. 2003
[31] B. G. Lee. A new algorithm to compute the discrete cosine transform. IEEE Transactions ASSP-32. 1984, (6)
[32] T. D. Tran. The binDCT: fast multiplierless approximation of the DCT. IEEE Signal Processing Letters. Jun 2000, 7(6): 141-144
[33] F. W. Wheeler, W. A. Pearlman. SPIHT image compression without lists. Proc. of the International Conf. on Acoustics, Speech and Signal Processing. Vancouver, Canada, 2000, 2047-2050
[34] F. W. Wheeler, W. A. Pearlman. Low memory packetized spiht image compression. M. B. Mattthews, (Editor) Conf. Record of The Thirty-Third Asilomar Conf. on Signals, Systems and Computers(M. B. Matthews, ed.). Pacific Grove, CA, 1999, vol. 2, 1193-1197
[35] C. Y. Su, B. F. Wu. A low memory zerotree coding for arbitrarily shaped objects. IEEETransactions on Image Processing. March 2003, 12(3): 271-282
[36] Yong Sun, Hui Zhang, Guangshu Hu. Real-time implementation of a new low-memory SPIHT image coding algorithm using dsp chip. IEEE Transactions on Image Processing. September 2002, 11 (9): 1112-1115
[37] J. M. Shapiro. A fast technique for identifying zerotrees in the EZW algorithm. Proc. of the International Conf. on Acoustics, Speech and Signal Processing. 1996, 1455-1458
[38] Zhao Debin, Gao Wen. Low complexity and low memory entropy coder for image compression. IEEE Transactions on Circuit and System for Video Technology. 2001, 11 (10):1140-1145
[39] Aaron Deever, Sheila Hemami. Efficient sign coding and estimation of zero-quantized coefficients in embedded wavelet image codecs. IEEE Transactions Image Processing. 2003, 12(4):420-429
[40] A. Said, W. A. Pearlman. SPIHT image compression home page. Online, 1996. www.cipr.rpi.edu/research/SPIHT/
[41] 成礼智，王洪霞，罗永．小波的理论与应用．研究生教材丛书，科学出版社，2003
[42] 曾文曲，文有为，孙炜．分形小波与图像压缩．东北大学出版社，2002
[43] 周娟，罗建书，乔士东．结合小波变换的零搜索分型图像编码．中国图像图形学报．July 2001，6(7)
[44] 冯晋臣，赵荣堂．陆地卫星CCT数据压缩处理研究．南京林产工业学院学报．1981．3：98-114
[45] 张杨．空间太阳望远镜中图像压缩单元软件设计．山东科技大学学报．2003，22(1)
[46] 钟是等．小波变换在遥感图像压缩中应用研究．中国图形图像学报．October 1998，3(10)
[47] 郭武，梅丽，罗建书．一种基于SPIHT的ROI图形编码．中国空间科学技术．Feb 2003(1)