Advertisement

Figure-ground image segmentation using feature-based multi-objective genetic programming techniques

  • Yuyu LiangEmail author
  • Mengjie Zhang
  • Will N. Browne
Original Article
  • 100 Downloads

Abstract

Figure-ground image segmentation is a process of separating regions of interest from a target image. Genetic programming has been employed to evolve segmentors that have the potential to capture high variations of images and conduct accurate segmentation. However, GP-based methods tend to evolve complex segmentors that have large sizes, are computationally expensive and difficult to interpret. Therefore, this work aims to balance the solution functionality with the complexity by applying multi-objective techniques to GP. Specifically, NSGA-II (nondominated sorting genetic algorithm) and SPEA2 (strength Pareto evolutionary algorithm) are selected as the base multi-objective techniques, in which a new Pareto dominance mechanism is designed, thus creating two new multi-objective techniques—INSGA-II (improved NSGA-II) and ISPEA2 (improved SPEA2). By applying the INSGA-II and ISPEA2 to GP, respectively, two novel multi-objective GP (MOGP) methods are proposed—INSGP and ISPGP. Both methods have two objectives: a solution functionality measure (i.e. the classification accuracy) and a solution complexity measure based on an exponential function. The results show that the proposed MOGP methods can evolve solutions with good trade-offs between the functionality and complexity, and INSGP is better at keeping solution diversity than ISPGP for the segmentation tasks in this paper. Moreover, the analyses on the evolved segmentors show that certain discriminatory patterns can be captured.

Keywords

Figure-ground image segmentation Genetic programming Multi-objective optimization Solution complexity 

Notes

Compliance with ethical standards

Conflict of interest

Author Yuyu Liang holds the doctoral scholarship from China Scholarship Council. Author Mengjie Zhang is the chair of IEEE Computational Intelligence Society, and a committee member of the IEEE New Zealand Central Section. Author Will N. Browne was editor-in-chief for the Australasian Conference on Robotics and Automation 2012, and is a member of the ACM SIGEVO group.

References

  1. 1.
    Al-Sahaf H, Song A, Neshatian K, Zhang M (2012) Extracting image features for classification by two-tier genetic programming. In: 2012 IEEE Congress on Evolutionary Computation (CEC), pp 1–8. IEEEGoogle Scholar
  2. 2.
    Bleuler S, Brack M, Thiele L, Zitzler E (2001) Multiobjective genetic programming: Reducing bloat using SPEA2. In: Proceedings of the 2001 Congress on Evolutionary Computation, vol 1. IEEE, pp 536–543Google Scholar
  3. 3.
    Borenstein E, Ullman S (2002) Class-specific, top-down segmentation. In: ECCV 2002. Springer, pp 109–122Google Scholar
  4. 4.
    Borenstein E, Ullman S (2004) Learning to segment. In: ECCV 2004. Springer, pp 315–328Google Scholar
  5. 5.
    Borenstein E, Ullman S (2008) Combined top-down/bottom-up segmentation. IEEE Trans Pattern Anal Mach Intell 30(12):2109–2125CrossRefGoogle Scholar
  6. 6.
    Branke J (2008) Consideration of partial user preferences in evolutionary multiobjective optimization. In: Multiobjective optimization. Springer, pp 157–178Google Scholar
  7. 7.
  8. 8.
    Davidson J, Savic D, Walters G (1999) Method for the identification of explicit polynomial formulae for the friction in turbulent pipe flow. J Hydroinform 1:115–126Google Scholar
  9. 9.
    De Jong ED, Watson RA, Pollack JB (2001) Reducing bloat and promoting diversity using multi-objective methods. In: Proceedings of the 3rd annual conference on genetic and evolutionary computation. Morgan Kaufmann Publishers, pp 11–18Google Scholar
  10. 10.
    Deb K (2015) Multi-objective evolutionary algorithms. In: Kacprzyk J, Pedrycz W (eds) Springer handbook of computational intelligence. Springer, Berlin, pp 995–1015CrossRefGoogle Scholar
  11. 11.
    Deb K, Pratap A, Agarwal S, Meyarivan T (2002) A fast and elitist multiobjective genetic algorithm: NSGA-II. IEEE Trans Evol Comput 6(2):182–197CrossRefGoogle Scholar
  12. 12.
    Everingham M, Eslami SA, Van Gool L, Williams CK, Winn J, Zisserman A (2014) The PASCAL visual object classes challenge: a retrospective. Int J Comput Vis 111(1):98–136CrossRefGoogle Scholar
  13. 13.
    Gill G, Toews M, Beichel RR (2014) Robust initialization of active shape models for lung segmentation in CT scans: a feature-based atlas approach. J Biomed Imaging 2014:13Google Scholar
  14. 14.
    Khan W (2013) Image segmentation techniques: a survey. J Image Graph 1(4):166–170Google Scholar
  15. 15.
    Koza JR. What is genetic programming (GP)? http://www.genetic-programming.com/
  16. 16.
    Koza JR (1992) Genetic programming: on the programming of computers by natural selection. MIT Press, CambridgezbMATHGoogle Scholar
  17. 17.
    Liang Y, Zhang M, Browne WN (2015) A supervised figure-ground segmentation method using genetic programming. In: Applications of evolutionary computation. Springer, pp 491–503Google Scholar
  18. 18.
    Liang Y, Zhang M, Browne WN (2016) Multi-objective genetic programming for figure-ground image segmentation. In: Artificial life and computational intelligence. Springer, pp 134–146Google Scholar
  19. 19.
    Liang Y, Zhang M, Browne WN (2017) Genetic programming for evolving figure-ground segmentors from multiple features. Appl Soft Comput 51:83–95CrossRefGoogle Scholar
  20. 20.
    Liu CY, Iglesias JE, Tu Z, Initiative ADN et al (2013) Deformable templates guided discriminative models for robust 3D brain MRI segmentation. Neuroinformatics 11(4):447–468CrossRefGoogle Scholar
  21. 21.
    Liu J, Wang J (2014) Application of snake model in medical image segmentation. J Converg Inf Technol 9(1):105–109Google Scholar
  22. 22.
    Lizárraga GL, Rionda SB (2009) On the diversity of non-dominated sets. http://www.micai.org/2009/proceedings/complementary/cd/ws-imso/191/diversity.pdf. Accessed 08 Nov 2017
  23. 23.
    Luke S, Panait L (2006) A comparison of bloat control methods for genetic programming. Evol Comput 14(3):309–344CrossRefGoogle Scholar
  24. 24.
    McKnight PE, Najab J (2010) Mann–Whitney U test. Corsini Encyclopedia of PsychologyGoogle Scholar
  25. 25.
    Poli R (1996) Genetic programming for feature detection and image segmentation. In: Evolutionary computing. Springer, pp 110–125Google Scholar
  26. 26.
    Poli R (2003) A simple but theoretically-motivated method to control bloat in genetic programming. In: Genetic programming. Springer, pp 204–217Google Scholar
  27. 27.
    Poli R, Langdon WB, McPhee NF, Koza JR (2008) A field guide to genetic programming. Lulu.comGoogle Scholar
  28. 28.
    Sarro F, Ferrucci F, Gravino C (2012) Single and multi objective genetic programming for software development effort estimation. In: Proceedings of the 27th annual ACM symposium on applied computing. ACM, pp 1221–1226Google Scholar
  29. 29.
    Sasaki Y et al (2007) The truth of the f-measure. Teaching and tutorial materials 1(5)Google Scholar
  30. 30.
    Segura C, Coello CAC, Miranda G, León C (2016) Using multi-objective evolutionary algorithms for single-objective constrained and unconstrained optimization. Ann Oper Res 240(1):217–250CrossRefzbMATHMathSciNetGoogle Scholar
  31. 31.
    Shao L, Liu L, Li X (2014) Feature learning for image classification via multiobjective genetic programming. IEEE Trans Neural Netw Learn Syst 25(7):1359–1371CrossRefGoogle Scholar
  32. 32.
    Singh T, Kharma N, Daoud M, Ward R (2009) Genetic programming based image segmentation with applications to biomedical object detection. In: Proceedings of the 11th annual conference on genetic and evolutionary computation. ACM, pp 1123–1130Google Scholar
  33. 33.
    Song A, Ciesielski V (2008) Texture segmentation by genetic programming. Evol Comput 16(4):461–481CrossRefGoogle Scholar
  34. 34.
    Soule T, Foster JA (1998) Effects of code growth and parsimony pressure on populations in genetic programming. Evol Comput 6(4):293–309CrossRefGoogle Scholar
  35. 35.
    Thada V, Jaglan V (2013) Comparison of jaccard, dice, cosine similarity coefficient to find best fitness value for web retrieved documents using genetic algorithm. Int J Innov Eng Technol 2(4):202–205Google Scholar
  36. 36.
    Wang B, Singh HK, Ray T (2015) A multi-objective genetic programming approach to uncover explicit and implicit equations from data. In: 2015 IEEE Congress on Evolutionary Computation (CEC). IEEE, pp 1129–1136Google Scholar
  37. 37.
    Yeghiazaryan V, Voiculescu I (2015) An overview of current evaluation methods used in medical image segmentation. Technical report, CS-RR-15-08, Department of Computer Science, University of Oxford, Oxford, UKGoogle Scholar
  38. 38.
    Zitzler E, Brockhoff D, Thiele L (2007) The hypervolume indicator revisited: On the design of Pareto-compliant indicators via weighted integration. In: Evolutionary multi-criterion optimization. Springer, pp 862–876Google Scholar
  39. 39.
    Zitzler E, Laumanns M, Thiele L, Zitzler E, Zitzler E, Thiele L, Thiele L (2001) SPEA2: improving the strength Pareto evolutionary algorithmGoogle Scholar
  40. 40.
    Zou W, Bai C, Kpalma K, Ronsin J (2014) Online global transfer for automatic figure-ground segmentation. IEEE Trans Image Process 23(5):2109–2121CrossRefMathSciNetGoogle Scholar

Copyright information

© The Natural Computing Applications Forum 2017

Authors and Affiliations

  1. 1.School of Engineering and Computer ScienceVictoria University of WellingtonWellingtonNew Zealand

Personalised recommendations