Abstract
In image segmentation tasks, shadow, clutter background and various interference factors in the image increase the difficulty of segmentation, and lead to unsatisfied results. To solve these problems, this paper proposes an image segmentation algorithm combining MRF (Markov Random Field) with deep learning shape priori. The target shape priori information is modelled and generated from deep learning models: Restricted Boltzmann Machine (RBM), Deep Belief Network (DBN), and Deep the Boltzmann Machine (DBM). Then the shape priori information is further defined and added to the energy function of the MRF image segmentation algorithm. Since the shape priori restricts the target shape and restrains the interference factors, better segmentation results are obtained. The proposed method is compared with traditional MRF and some comparable image segmentation methods in two datasets, the experiments results demonstrate the effective of the proposed method.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Belongie, S., Carson, C., Greenspan, H., et al.: Color-and texture-based image segmentation using EM and its application to content-based image retrieval. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 675–682 (1998)
Chen, J., Pappas, T.N., Mojsilovic, A., et al.: Adaptive image segmentation based on color and texture. In: Proceedings of the IEEE International Conference on Image Processing, pp. 777–780 (2002)
LeCun, Y., Bengio, Y., Hinton, G.E.: Deep learning. Nature 521(7553), 436–444 (2015)
Juan, Z., Xili, W., Jiangong, Y.: Shape modeling method based on deep learning. Chin. J. Comput. 41(01), 132–144 (2018)
Hinton, G.E., Osindero, S., Teh, Y.-W.: A fast learning algorithm for deep belief nets. Neural Comput. 18(7), 1527–1554 (2006)
Salakhutdinov, R., Hinton, G.E.: Deep Boltzmann machines. J. Mach. Learn. Res. 5(2), 1967–2006 (2009)
Hinton, G.E.: Training products of experts by minimizing contrastive divergence. Neural Comput. 14(8), 1771–1800 (2002)
Huang, L., Nie, J., Wei, Z.: Human body segmentation based on shape constraint. Mach. Vis. Appl. 2017(2), 1–10 (2017)
Hinton, G.E.: A practical guide to training restricted Boltzmann machines. Momentum 9(1), 599–619 (2010)
Boykov, Y., Kolmogorov, V.: An experimental comparison of min-cut/max-flow algorithms for energy minimization in vision. IEEE Trans. Pattern Anal. Mach. Intell. 26(9), 1124–1137 (2004)
Zhang, W.: Markov random field based object segmentation combining edge and shape prior. J. Chongqing Univ. Technol. 10, 79–85 (2014)
Zhang, W.: Research on Image Classification Based on Probability Graph Model. ShaanXi Normal University, Xi’an (2013)
Borenstein, E., Sharon, E., Ullman, S.: Combining top-down and bottom-up segmentation. In: CVPR (2004)
Welinder, P., Branson, S., Mita, T., Wah, C., Schroff, F., Belongie, S., Perona, P.: Caltech-UCSD Birds 200. Technical report CNS-TR-2010-001, California Institute of Technology (2010)
Fischer, A., Igel, C.: Training restricted Boltzmann machines: an introduction. Patt. Recogn. 47(1), 25–39 (2014)
Elfwing, S., Uchibe, E., Doya, K.: Expected energy-based restricted Boltzmann machine for classification. Neural Netw. 64, 29–38 (2015)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer International Publishing AG, part of Springer Nature
About this paper
Cite this paper
Wang, Y., Wang, X. (2018). Image Segmentation Based on MRF Combining with Deep Learning Shape Priors. In: Geng, X., Kang, BH. (eds) PRICAI 2018: Trends in Artificial Intelligence. PRICAI 2018. Lecture Notes in Computer Science(), vol 11013. Springer, Cham. https://doi.org/10.1007/978-3-319-97310-4_21
Download citation
DOI: https://doi.org/10.1007/978-3-319-97310-4_21
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-97309-8
Online ISBN: 978-3-319-97310-4
eBook Packages: Computer ScienceComputer Science (R0)