劉新武??
摘 要 為了快速地去除圖像中的泊松噪聲, 本文在傳統(tǒng)的交替方向算法基礎(chǔ)上, 結(jié)合松弛算法提出了一個改進的快速交替最小化算法. 與經(jīng)典的數(shù)值算法相比,數(shù)值試驗表明提出的新算法不但能有效地實現(xiàn)泊松化圖像復(fù)原, 還能大幅度地提高數(shù)值計算的速率, 并顯著地減少電腦的CPU運行時間.
關(guān)鍵詞 圖像復(fù)原;泊松噪聲;全變差;交替最小化算法
中圖分類號 TP391 文獻標識碼 A
Alternating Minimization Algorithm
for Poissonian Image Restoration
LIU Xinwu
(School of Mathematics and Computational Science, Hunan University of Science
and Technology, Xiangtan, Hunan 411201,China)
Abstract To quickly remove Poisson noise, based on the traditional alternating direction method, this paper combined the relaxation method and proposed an improved alternating minimization algorithm. Compared with the classical numerical algorithm, numerical simulations demonstrate that the proposed strategy not only removes Poisson noise efficiently, but improves the speed of calculation substantially and reduces the computer CPU time noticeably.
Key words image restoration; Poisson noise; total variation; alternating minimization algorithm
1 引 言
圖像在形成、傳輸和存儲過程中, 不可避免地會受到噪聲的影響. 譬如, 在天文成像[1]和電子顯微鏡成像[2,3]中, 獲得的圖像就往往會受到泊松噪聲污染,并出現(xiàn)明顯的降質(zhì)現(xiàn)象,因此圖像復(fù)原就顯得尤為重要. 目前,圖像復(fù)原技術(shù)已在天文學(xué)、醫(yī)學(xué)、刑偵、軍事以及金融學(xué)等領(lǐng)域得到了廣泛的應(yīng)用. 例如,在商業(yè)和金融行業(yè)中,一個新興的融合信息科學(xué)、金融學(xué)和管理學(xué)的先進金融信息技術(shù)(如模式識別、人工智能等)已有效地應(yīng)用于金融票據(jù)識別、金融票據(jù)影像處理及打印中,并成功地解決了一系列經(jīng)濟領(lǐng)域中的熱點和難點問題.
4 結(jié) 論
本文研究了一個基于TV正則化模型的泊松化圖像復(fù)原問題. 為了提高數(shù)值計算的速率, 本文結(jié)合傳統(tǒng)的交替方向法和松弛算法, 提出了一個改進的交替最小化算法. 數(shù)值試驗表明, 新算法在泊松去噪中具有顯著的優(yōu)越性和高效性.同時,該算法也必將在金融行業(yè)中的金融票據(jù)識別和票據(jù)影像處理中得到進一步的發(fā)展和應(yīng)用.
參考文獻
[1] E BRATSOLIS, M SIGELLA. A spatial regularization method preserving local photometry for RichardsonLucy restoration [J]. Astronomy Astrophysics, 2001, 375(3): 1120-1128.
[2] N DAY, L BLANCFERAUD, C ZIMMER, et al. RichardsonLucy algorithm with total variation regularization for 3D confocal microscope deconvolution [J]. Microscopy Research and Technique, 2006, 69(4): 260-266.
[3] A KRYVANOS, J HESSER, G STEIDL. Nonlinear image restoration methods for marker extraction in 3D fluorescent microscopy[C]//Proceedings of SPIE, San Jose, CA, USA. Computational Imaging III, 2005, 5674: 432-443.
[4] T LE, R CHARTRAND, T ASAKI. A variational approach to constructing images corrupted by Poisson noise [J]. Journal of Mathematical Imaging and Vision, 2007, 27(3): 257-263.
[5] A SAWATZKY, C BRUNE, F WUBBELING, T KOSTERS, K SCHAFERS, M BURGER. Accurate EMTV algorithm in PET with low SNR[C]//IEEE Nuclear Science Symposium Conference Record, Dresden, Germany, 2008, 5133-5137.
[6] C BRUNE, A SAWATZKY, M BURGER. Primal and dual Bregman methods with application to optical nanoscopy [J]. International Journal of Computer Vision, 2011, 92(2): 211-229.
[7] M FIGUEIREDO, J BIOUCASDIAS. Restoration of Poissonian images using alternating direction optimization [J]. IEEE Transactions on Image Processing, 2010, 19(12): 3133-3145.
[8] S SETZER, G STEIDL, T TEUBER. Deblurring Poissonian images by split Bregman techniques [J]. Journal of Visual Communication and Image Representation, 2010, 21(3): 193-199.
[9] X LIU, L HUANG. Total bounded variationbased Poissonian images recovery by split Bregman iteration [J]. Mathematical Methods in the Applied Sciences, 2012, 35(5): 520-529.
[10]T F CHAN, P MULET. On the convergence of the lagged diffusivity fixed point method in total variation image restoration [J]. SIAM Journal on Numerical Analysis, 1999, 36(2): 354-367.
[11]C VOGEL, M OMAN. Iteration methods for total variation denoising [J]. SIAM Journal on Scientific Computing, 1996, 17(1): 227-238.
[12]A CHAMBOLLE. An algorithm for total variation minimization and application [J]. Journal of Mathematical Imaging and Vision, 2004, 20(1-2): 89-97.
[13]M K NG, L QI, Y YANG, Y HUANG. On semismooth Newtons methods for total variation minimization [J]. Journal of Mathematical Imaging and Vision, 2007, 27(3): 265-276.
[14]T GOLDSTEIN, S OSHER. The split Bregman algorithm for L1 regularized problems [J]. SIAM Journal on Imaging Sciences, 2009, 2(2): 323-343.
[15]J F CAI, S OSHER, Z SHEN. Split Bregman methods and frame based image restoration [J]. Multiscale Modeling & Simulation, 2009, 8(2): 337-369.
[16]X LIU, L HUANG. Split Bregman iteration algorithm for total bounded variation regularization based image deblurring [J]. Journal of Mathematical Analysis and Applications, 2010, 372(2): 486-495.
[17]W YIN, S OSHER, D GOLDFARB, J DARBON. Bregman iterative algorithms for L1minimization with applications to compressed sensing [J]. SIAM Journal on Imaging Sciences, 2008, 1(1): 143-168.
[18]RQ JIA, H ZHAO, W ZHAO. Relaxation methods for image denoising based on difference schemes [J]. Multiscale Modeling & Simulation, 2011, 9(1): 355-372.
[7] M FIGUEIREDO, J BIOUCASDIAS. Restoration of Poissonian images using alternating direction optimization [J]. IEEE Transactions on Image Processing, 2010, 19(12): 3133-3145.
[8] S SETZER, G STEIDL, T TEUBER. Deblurring Poissonian images by split Bregman techniques [J]. Journal of Visual Communication and Image Representation, 2010, 21(3): 193-199.
[9] X LIU, L HUANG. Total bounded variationbased Poissonian images recovery by split Bregman iteration [J]. Mathematical Methods in the Applied Sciences, 2012, 35(5): 520-529.
[10]T F CHAN, P MULET. On the convergence of the lagged diffusivity fixed point method in total variation image restoration [J]. SIAM Journal on Numerical Analysis, 1999, 36(2): 354-367.
[11]C VOGEL, M OMAN. Iteration methods for total variation denoising [J]. SIAM Journal on Scientific Computing, 1996, 17(1): 227-238.
[12]A CHAMBOLLE. An algorithm for total variation minimization and application [J]. Journal of Mathematical Imaging and Vision, 2004, 20(1-2): 89-97.
[13]M K NG, L QI, Y YANG, Y HUANG. On semismooth Newtons methods for total variation minimization [J]. Journal of Mathematical Imaging and Vision, 2007, 27(3): 265-276.
[14]T GOLDSTEIN, S OSHER. The split Bregman algorithm for L1 regularized problems [J]. SIAM Journal on Imaging Sciences, 2009, 2(2): 323-343.
[15]J F CAI, S OSHER, Z SHEN. Split Bregman methods and frame based image restoration [J]. Multiscale Modeling & Simulation, 2009, 8(2): 337-369.
[16]X LIU, L HUANG. Split Bregman iteration algorithm for total bounded variation regularization based image deblurring [J]. Journal of Mathematical Analysis and Applications, 2010, 372(2): 486-495.
[17]W YIN, S OSHER, D GOLDFARB, J DARBON. Bregman iterative algorithms for L1minimization with applications to compressed sensing [J]. SIAM Journal on Imaging Sciences, 2008, 1(1): 143-168.
[18]RQ JIA, H ZHAO, W ZHAO. Relaxation methods for image denoising based on difference schemes [J]. Multiscale Modeling & Simulation, 2011, 9(1): 355-372.
[7] M FIGUEIREDO, J BIOUCASDIAS. Restoration of Poissonian images using alternating direction optimization [J]. IEEE Transactions on Image Processing, 2010, 19(12): 3133-3145.
[8] S SETZER, G STEIDL, T TEUBER. Deblurring Poissonian images by split Bregman techniques [J]. Journal of Visual Communication and Image Representation, 2010, 21(3): 193-199.
[9] X LIU, L HUANG. Total bounded variationbased Poissonian images recovery by split Bregman iteration [J]. Mathematical Methods in the Applied Sciences, 2012, 35(5): 520-529.
[10]T F CHAN, P MULET. On the convergence of the lagged diffusivity fixed point method in total variation image restoration [J]. SIAM Journal on Numerical Analysis, 1999, 36(2): 354-367.
[11]C VOGEL, M OMAN. Iteration methods for total variation denoising [J]. SIAM Journal on Scientific Computing, 1996, 17(1): 227-238.
[12]A CHAMBOLLE. An algorithm for total variation minimization and application [J]. Journal of Mathematical Imaging and Vision, 2004, 20(1-2): 89-97.
[13]M K NG, L QI, Y YANG, Y HUANG. On semismooth Newtons methods for total variation minimization [J]. Journal of Mathematical Imaging and Vision, 2007, 27(3): 265-276.
[14]T GOLDSTEIN, S OSHER. The split Bregman algorithm for L1 regularized problems [J]. SIAM Journal on Imaging Sciences, 2009, 2(2): 323-343.
[15]J F CAI, S OSHER, Z SHEN. Split Bregman methods and frame based image restoration [J]. Multiscale Modeling & Simulation, 2009, 8(2): 337-369.
[16]X LIU, L HUANG. Split Bregman iteration algorithm for total bounded variation regularization based image deblurring [J]. Journal of Mathematical Analysis and Applications, 2010, 372(2): 486-495.
[17]W YIN, S OSHER, D GOLDFARB, J DARBON. Bregman iterative algorithms for L1minimization with applications to compressed sensing [J]. SIAM Journal on Imaging Sciences, 2008, 1(1): 143-168.
[18]RQ JIA, H ZHAO, W ZHAO. Relaxation methods for image denoising based on difference schemes [J]. Multiscale Modeling & Simulation, 2011, 9(1): 355-372.