Influence of Clique Potential Parameters on Classification Using Bayesian MRF Model for Remote Sensing Image in Dali Erhai Basin

Abstract:

Image classification of remote sensing data is an important topic and long-term tasks in applications [1]. Markov random field (MRF) has more advantages in processing contextual information [2]. Bayesian approach enables the incorporation of prior model and likelihood distribution, this paper has formulated a Bayesian-MRF classification model based on MAP-ICM framework. It uses Potts model in label field and assume Gaussian distribution in observation field. According to maximum a posteriori (MAP) criterion, each new classified label can be obtained by the minimum of energy using Iterated Conditional Modes (ICM) algorithm. Finally, classification tasks are carried out by Bayesian-MRF classification model. Experimental results show that: (1) Clique potential parameters affect classification greatly. When it is 0.5, the classification accuracy reaches maximum with the best classification result for study area of Dali Erhai Lake basin using landsat TM data. (2) Bayesian MRF model have obvious advantages in classification for neighbourhood pixels so that it can separate Shadow class from Water class because the Shadow in mountain areas is very similar to Water in spectrum. In this case study, the best classification accuracy reaches 95.8%. The approaches and results will have important reference value for applications such as land use/cover classification, environment/ecological monitoring etc.

• Periodical: Advanced Materials Research (Volume 658)
• Pages: 508-512
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.658.508
• Online since: January 2013
• Authors: Hai Jun Duan, Guang Min Wu, Dan Liu, John D. Mai, Jian Ming Chen
• Keywords: Bayesian-MRF Classification Model, Clique Potential Parameter, Iterated Conditional Modes (ICM), Landsat TM, Markov Random Field (MRF), Maximum a Posterior (MAP)

References

[1] B. Charles and S. Ken, A Generalized Gaussian Image Model for Edge-Preserving MAP Estimation, IEEE Transactions on Image Processing, vol. 2, no. 3, p.296–310, July, (1993).

DOI: 10.1109/83.236536

[2] Dina E. Melas and Simon P. Wilson, Double Markov Random Fields and Bayesian Image Segmentation, IEEE Transactions on Image Processing, vol. 50, no. 2, p.357–365, Feb. (2002).

DOI: 10.1109/78.978390

[3] B. Marc, K. Zoltan, Y. Shan, Z. Josiane, Bayesian image classification using Markov random fields, Image and Vision Computing, vol. 14, pp.285-295, (1996).

DOI: 10.1016/0262-8856(95)01072-6

[4] S. Krishnamachari and R. Chellappa, Multiresolution Gauss-Markov random field models for texture segmentation, IEEE Transactions on Image Processing, vol. 6, no. 2, p.251–267, Feb. (1997).

DOI: 10.1109/83.551696

[5] K. Koray, E. K. Ercan, and S. Bulent, Bayesian Separation of Images Modeled With MRFs Using MCMC, IEEE Transactions on Image Processing, vol. 18, no. 5, May (2009).

DOI: 10.1109/tip.2009.2012905

[6] Stan Z. Li. Markov Random Field Modeling in Image Analysis (Third Edition). London: Springer-Verlag, (2009).

[7] AK. Jain and SG. Nadabar, MRF Model-Based Segmentation of Range Images, IEEE Conference on Computer Vision, pp.667-671, (1990).

DOI: 10.1109/iccv.1990.139615

[8] L. Wang, J. Liu, Texture classification using multiresolution Markov random field models, Pattern Recognition Letters, vol. 20, pp.171-182, (1999).

DOI: 10.1016/s0167-8655(98)00129-9

[9] M. Schroder, H. Rehrauer, K. Seidel, and M. Datcu, Spatial information retrieval from remote-sensing images-Part II: Gibbs-Markov random fields, IEEE Transactions on Geoscience and Remote Sensing, vol. 36, no. 5, pp.1416-1455, Sep. (1998).

DOI: 10.1109/36.718848

[10] B. Tso and R. C. Olsen, Scene classification using combined spectral, textural and contextual information, in Proceedings SPIE, vol. 5425, pp.135-146.

[11] Y. Hu and T. J. Dennis, Simulated annealing and iterated conditional modes with selective and confidence enhanced update schemes, in Proc. 5th Annu. IEEE Symp. Comput. -Based Med. Syst., pp.257-264, (1992).

DOI: 10.1109/cbms.1992.244940

[12] Robert A. Weisenseel and W. Clem Karl, Markov random field segmentation methods for SAR target chips, in Proceedings SPIE, vol. 3721, pp.145-156, (1999).

[13] S. Krishnamachari and R. Chellappa, Multiresolution Gauss-Markov random field models for texture segmentation, IEEE Transactions on Image Processing, vol. 6, no. 2, p.251–267, Feb. (1997).

DOI: 10.1109/83.551696

[14] Max Mignotte, A Label Field Fusion Bayesian Model and Its Penalized Maximum Rand Estimator for Image Segmentation, IEEE Transactions on Image Processing, vol. 19, no. 6, p.1610–1624, Jun. (2010).

DOI: 10.1109/tip.2010.2044965