Publicado
2021-01-01

Recuperación de imágenes usando modelos auto-regresivos condicionales: CAR e IAR

Image recovery using conditional autoregressive models: CAR and IAR

DOI: https://doi.org/10.15332/23393076.6376
Danna Lesley Cruz Reyes

Resumen (es)

Este artículo realiza la estimación Bayesiana de campos aleatorios gausianos de Markov. En particular, se propone realizar un análisis de dependencia espacial por medio de un grafo que caracteriza las intensidades observadas de una imagen con un modelo ampliamente utilizado en estadística espacial y geoestadística conocido como modelo autorregresivo condicional (CAR por sus siglas en inglés). Este modelo es útil para obtener distribuciones conjuntas multivariadas de un vector aleatorio basado en especificaciones condicionales univariadas. Estas especificaciones condicionales se basan en las propiedades de Markov, de modo que la distribución condicional de un componente del vector aleatorio depende solo de un conjunto de vecinos, definido por el grafo. Los modelos autorregresivos condicionales son casos particulares de campos aleatorios de Markov y se utilizan como distribuciones \textit{a priori}, que combinadas con la información contenida en los datos de la muestra (función de verosimilitud), inducen una distribución \textit{a posteriori} en las que se basa la estimación. El modelo CAR tiene un caso particular llamado IAR, en el cual, la distribución \textit{a priori} no es propia. En este artículo se aplica ambos modelos haciendo una comparación entre ellos. Todos los parámetros del modelo se estiman en un entorno completamente Bayesiano, utilizando el algoritmo Metropolis-Hastings. Los procedimientos completos de estimación posterior se ilustran y comparan utilizando varios ejemplos artificiales. Para estos experimentos, el modelo CAR y el modelo IAR se comporta muy favorablemente con imágenes homogéneas

Palabras clave (es): Procesamiento probabilístico de imágenes, restauración de imágenes, modelado Bayesiano, modelo autorregresivo condicional

Resumen (en)

This article performs Bayesian estimation of Gaussian Markov random fields. In particular, it is proposed to perform a spatial dependency analysis by means of a graph that characterizes the observed intensities of an image with a model widely used in spatial statistics and geostatistics known as the conditional autoregressive model (CAR). This model is useful for obtaining multivariate joint distributions from a random vector based on univariate conditional specifications. These conditional specifications are based on the Markov properties, so that the conditional distribution of a component of the random vector depends only on a set of neighbors, defined by the graph. Conditional autoregressive models are particular cases of random Markov fields and are used as \textit{a priori} distributions, which, combined with the information contained in the sample data (likelihood function), induce a \textit{a posteriori} distribution on which the estimate is based. The CAR model has a particular case called IAR, in which the \textit{a priori} distribution is not proper, in this article both models are applied making a comparison between them. All model parameters are estimated in a completely Bayesian environment, using the Metropolis-Hastings algorithm. The complete estimation procedures are illustrated and compared using various artificial examples. For these experiments, the CAR model and the IAR model performed very favorably with homogeneous images.

Palabras clave (en): Probabilistic image processing, image restoration, Bayesian modeling, conditional autoregressive model

Referencias

Arslan, O. & Akyurek, (2018), `Spatial modelling of air pollution from pm10 and so2 concentrations during winter season in marmara region (2013-2014)',International Journal of Environment and Geoinformatics pp. 1 -16.

Assun_c~ao, R. & Krainski, E. (2009), `Neighborhood dependence in bayesian spatial models', Biometrical Journal 51(5), 851-869.*https://onlinelibrary.wiley.com/doi/abs/10.1002/bimj.200900056

Banerjee, S., P. Carlin, B. & E. Gelfand, A. (2004), `Hierarchical modeling and analysis of spatial data', Chapman and Hall/CRC Monographs on Statistical and Applied Probability; 101.

Besag, J. (1974), `Spatial interaction and the statistical analysis of lattice systems', Journal of the Royal Statistical Society. Series B (Methodological) 36(2), 192-236.*http://www.jstor.org/stable/2984812

Besag, J. (1986), `On the statistical analysis of dirty pictures', Journal of the Royal Statistical Society. Series B (Methodological) 48, 259-302.

Besag, J. & Higdon, D. (1999), `Bayesian analysis of agricultural _eld experiments', Journal of the Royal Statistical Society: Series B (Statistical Methodology) 61(4), 691-746. *https://rss.onlinelibrary.wiley.com/doi/abs/10.1111/1467-9868.00201

Besag, J. & Kooperberg, C. (1995), `On conditional and intrinsic autoregressions', Biometrika 82(4), 733-746.*https://doi.org/10.1093/biomet/82.4.733

Chen, C.-C. & Huang, C.-L. (1993), `Markov random _elds for texture classi_cation', Pattern Recognition Letters 14(11), 907- 914. *http://www.sciencedirect.com/science/article/pii/0167865593901557

Cross, G. R. & Jain, A. K. (1983), `Markov random _eld texture models', IEEE Transactions on Pattern Analysis and Machine Intelligence PAMI-5, 25-39.

Elliott, P. & Wartenberg, D. E. (2004), `Spatial epidemiology: Current approaches and future challenges'.

Givens, G. H. & Hoeting, J. A. (2012), Computational statistics, 2 edn, John Wiley & Sons, Hoboken, NJ, USA. Halim, S. (2008), `Modi_ed ising model for generating binary images', Jurnal Informatika 8.

Horiguchi, T., Honda, Y. & Miya, M. (1997), `Restoration of digital images of the alphabet by using ising models', Physics Letters A 227(5), 319 -324. *http://www.sciencedirect.com/science/article/pii/S0375960197000807

LeSage, J. P. & Thomas-Agnan, C. (2015), `Interpreting spatial econometric origin-destination flow models', Journal of Regional Science 55(2), 188-208. *https://onlinelibrary.wiley.com/doi/abs/10.1111/jors.12114

Liu, S. & Cooper, D. B. (2010), `Ray markov random _elds for image-based 3d modeling: Model and e_cient inference', pp. 1530-1537.

Mao, J. & Jain, A. K. (1992), `Texture classi_cation and segmentation using multiresolution simultaneous autoregressive models', Pattern Recognition 25(2), 173-188.

Morris, R. D., Descombes, X. & Zerubia, J. (1996), `The ising/potts model is not well suited to segmentation tasks', pp. 263-266.

Qian, W. & Titterington, D. M. (1991), `Multidimensional markov chain models for image textures', Journal of the Royal Statistical Society: Series B (Methodological) 53(3), 661-674. *https://rss.onlinelibrary.wiley.com/doi/abs/10.1111/j.25176161.1991.tb01855.x

Rue, H. & Held, L. (2005), `Gaussian Markov random _elds: Theory and applications', 104. *http://dx.doi.org/10.1201/9780203492024

Van Leemput, K., Maes, F., Vandermeulen, D. & Suetens, P. (1999), `Automated model-based tissue classi_cation of mr images of the brain', IEEE Transactions on Medical Imaging 18(10), 897-908.

Ver Hoef, J., Peterson, E., Hooten, M., Hanks, E. & Fortin, M.-J. (2017), `Spatial autoregressive models for statistical inference from ecological data', Ecological Monographs .

Wall, M. M. (2004), `A close look at the spatial structure implied by the CAR and SAR models', Journal of Statistical Planning and Inference 121(2), 311-324. *http://www.sciencedirect.com/science/article/pii/S0378375803001113

Xing, W., Deng, N., Xin, B., Chen, Y. & Zhang, Z. (2019), `Investigation of a novel automatic micro image-based method for the recognition of animal _bers based on wavelet and markov random _eld', Micron 119, 88 -97. *http://www.sciencedirect.com/science/article/pii/S0968432818304797

Dimensions

PlumX

Visitas

461

Descargas

Los datos de descarga aún no están disponibles.

Cómo citar

Cruz Reyes, D. L. . (2021). Recuperación de imágenes usando modelos auto-regresivos condicionales: CAR e IAR. Comunicaciones En Estadística, 14(1), 1-14. https://doi.org/10.15332/23393076.6376