Abstract
Filter networks are a powerful tool for reducing image processing time and maintaining high image quality. They are composed of sparse sub-filters whose high sparsity ensures fast image processing. The filter network design is related to solving a sparse optimization problem where a cardinality constraint bounds below the sparsity level. In the case of sequentially connected sub-filters, which is the simplest network structure of those considered in this paper, a cardinality-constrained multilinear least-squares (MLLS) problem is to be solved. Even when disregarding the cardinality constraint, the MLLS is typically a large-scale problem characterized by a large number of local minimizers, each of which is singular and non-isolated. The cardinality constraint makes the problem even more difficult to solve. An approach for approximately solving the cardinality-constrained MLLS problem is presented. It is then applied to solving a bi-criteria optimization problem in which both the time and quality of image processing are optimized. The developed approach is extended to designing filter networks of a more general structure. Its efficiency is demonstrated by designing certain 2D and 3D filter networks. It is also compared with the existing approaches.
Similar content being viewed by others
References
Andersson M, Wiklund J, Knutsson H (1999) Filter networks. In: Namazi N (ed) Signal and image processing (SIP), Proceedings of the IASTED international conferences, October 18–21, 1999. IASTED/ACTA Press, Nassau, The Bahamas, pp 213–217
Andersson M, Burdakov O, Knutsson H, Zikrin S (2012) Global search strategies for solving multilinear least-squares problems. SQU J Sci 17(1):12–21
Ansari N, Hou E (2012) Computational intelligence for optimization. Springer Publishing Company Inc., Heidelberg
Baran T, Wei D, Oppenheim AV (2010) Linear programming algorithms for sparse filters design. IEEE Trans Signal Process 58(3):1605–1617
van den Berg E, Friedlander M (2009) Probing the Pareto frontier for basis pursuit solutions. SIAM J Sci Comput 31(2):890–912
Bigun J (2006) Vision and direction. Springer, Heidelberg
Blumensath T (2011) Sparsify. http://users.fmrib.ox.ac.uk/~tblumens/sparsify/sparsify.html. Accessed 1 May 2012
Blumensath T (2012) Accelerated iterative hard thresholding. Signal Process 92(3):752–756
Blumensath T, Davies ME (2009) Iterative hard thresholding for compressed sensing. Appl Comput Harmon Anal 27(3):265–274
Bracewell RN, Bracewell RN (1986) The Fourier transform and its applications, vol 31999. McGraw-Hill, New York
Candès E, Romberg J (2005) \(l_1\)-MAGIC: recovery of sparse signals via convex programming. Tech. rep, California Institute of Technology, Pasadena, CA
Chen S, Billings SA, Luo W (1989) Orthogonal least squares methods and their application to non-linear system identification. Int J Control 50(5):1873–1896
Donoho D, Stodden V, Tsaig Y (2007) Sparselab. http://sparselab.stan-ford.edu/. Accessed 1 May 2012
Dudgeon DE, Mersereau RM (1990) Multidimensional digital signal processing. Prentice Hall Professional Technical Reference, Englewood Cliffs
Ehrgott M (2005) Multicriteria optimization. Springer-Verlag New York Inc., Secaucus
Estepar RSJ (2007) Local structure tensor for multidimensional signal processing. Applications to Medical Image Analysis. Presses universitaires de Louvain, Louvain-la-Neuve, Belgium
Figueiredo MAT, Nowak RD, Wright SJ (2007) Gradient projection for sparse reconstruction: application to compressed sensing and other inverse problems. IEEE J Sel Top Signal Process 1(4):586–597
George J, Indu SP, Rajeev K (2008) Three dimensional ultrasound image enhancement using local structure tensor analysis. In: TENCON 2008–2008 IEEE region 10 conference, pp 1–6
Granlund G, Knutsson H (1995) Signal processing for computer vision. Kluwer, Dordrecht
Hemmendorff M, Andersson M, Kronander T, Knutsson H (2002) Phase-based multidimensional volume registration. IEEE Trans Med Imaging 21(12):1536–1543
Kim SJ, Koh K, Lustig M, Boyd S (2007) An efficient method for compressed sensing. Proc IEEE Int Conf Image Process 3:117–120
Knutsson H (1982) Filtering and reconstruction in image processing. Dissertation, Linköping University
Knutsson H, Andersson M, Wiklund J (1999) Advanced filter design. In: Proceedings of the 11th scandinavian conference on image analysis, Greenland, SCIA, pp 185–193
Knutsson H, Westin CF, Andersson M (2011) Representing local structure using tensors II. In: Heyden A, Kahl F (eds) Image analysis, vol 6688., Lecture notes in computer scienceSpringer, Berlin, pp 545–556
Langer M (2004) Design of fast multidimensional filters by genetic algorithms. Master thesis, Linköping University
Langer M, Svensson B, Brun A, Andersson M, Knutsson H (2005) Design of fast multidimensional filters using genetic algorithms. In: Rothlauf F, Branke J, Cagnoni S, Corne D, Drechsler R, Jin Y, Machado P, Marchiori E, Romero J, Smith G, Squillero G (eds) Applications of evolutionary computing, vol 3449., Lecture notes in computer scienceSpringer, Berlin, pp 366–375
Leardi R, Armanino C, Lanteri S, Alberotanza L (2000) Three-mode principal component analysis of monitoring data from venice lagoon. J Chemom 14(3):187–195
Leurgans S, Ross RT (1992) Multilinear models: applications in spectroscopy. Stat Sci 7(3):289–310
Lopes JA, Menezes JC (2003) Industrial fermentation end-product modelling with multilinear PLS. Chemom Intell Lab Syst 68(1):75–81
Lu WS, Hinamoto T (2011) Two-dimensional digital filters with sparse coefficients. Multidimens Syst Signal Process 22(1–3):173–189
Mietinen K (1999) Nonlinear multiobjective optimization. Kluwer Academic Publishers, Boston
Muthukrishnan S (2005) Data streams: algorithms and applications. Now Publishers, Boston
Nocedal J, Wright SJ (2006) Numerical optimization, 2nd edn. Springer, New York
Norell B, Burdakov O, Andersson M, Knutsson H (2011) Approximate spectral factorization for design of efficient sub-filter sequences. Tech. Rep. LiTH-MAT-R-2011-14, Department of Mathematics, Linköping University
Ortega JM, Rheinboldt WC (2000) Iterative solution of nonlinear equations in several variables, classics in applied mathematics, vol 30, reprint of the 1970 original edn. SIAM, Philadelphia, PA
Paatero P (1997) Least squares formulation of robust non-negative factor analysis. Chemom Intell Lab Syst 37(1):23–35
Pati YC, Rezaiifar R, Krishnaprasad P (1993) Orthogonal matching pursuit: recursive function approximation with applications to wavelet decomposition. In: Proceedings of the 27th annual asilomar conference on signals, systems, and computers vol 1, pp 40–44
Rusu C, Dumitrescu B (2012) Iterative reweighted \(l_1\) design of sparse FIR filters. Signal Process 92(4):905–911
Sun X, Zheng X, Li D (2013) Recent advances in mathematical programming with semi-continuous variables and cardinality constraint. J Oper Res Soc China 1(1):55–77
Svensson B, Andersson M, Knutsson H (2005a) Filter networks for efficient estimation of local 3-d structure. Proc IEEE Int Conf Image Process 3:573–576
Svensson B, Andersson M, Knutsson H (2005b) A graph representation of filter networks. In: Kalviainen H, Parkkinen J, Kaarna A (eds) Image analysis, vol 3540., Lecture notes in computer scienceSpringer, Berlin, pp 1086–1095
Tropp JA, Wright SJ (2010) Computational methods for sparse solution of linear inverse problems. Proc IEEE 98(6):948–958
Wang JH, Hopke PK, Hancewicz TM, Zhang SL (2003) Application of modified alternating least squares regression to spectroscopic image analysis. Anal Chim Acta 476(1):93–109
Westin CF, Wigström L, Loock T, Sjöqvist L, Kikinis R, Knutsson H (2001) Three-dimensional adaptive filtering in magnetic resonance angiography. J Magn Reson Imaging 14(1):63–71
Acknowledgments
This work was supported by the Swedish Research Council; the Linnaeus Center for Control, Autonomy, and Decision-making in Complex Systems (CADICS); the Swedish Foundation for Strategic Research (SSF) Strategic Research Center (MOVIII); the Linköping University Center for Industrial Information Technology (CENIIT).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Andersson, M., Burdakov, O., Knutsson, H. et al. Sparsity optimization in design of multidimensional filter networks. Optim Eng 16, 259–277 (2015). https://doi.org/10.1007/s11081-015-9280-3
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11081-015-9280-3