Bayesian Markov Chain Random Field Cosimulation for Improving Land Cover Classification Accuracy

This study introduces a Bayesian Markov chain random field (MCRF) cosimulation approach for improving land-use/land-cover (LULC) classification accuracy through integrating expert-interpreted data and pre-classified image data. The expert-interpreted data are used as conditioning sample data in cosimulation, and may be interpreted from various sources. The pre-classification can be performed using any convenient conventional method. The approach uses the recently suggested MCRF cosimulation algorithm (Co-MCSS) to take a pre-classified image as auxiliary data while performing cosimulations conditioned on expert-interpreted data. It was tested using a series of expert-interpreted data sets and an image data set pre-classified by the supervised maximum likelihood (SML) algorithm. Results show that with the density of the interpreted data (pixel labels) increasing from 0 to 1.81 % of total pixels, the accuracy of optimal classification maps from Co-MCSS improves by 8.49 to 20.96 %, being much higher than that generated by SML and those purely conditioned on expert-interpreted data. This means that expert-interpreted data may largely contribute to the accuracy of LULC classification from remotely sensed imagery, and inversely the pre-classified image data also can largely contribute to the accuracy of LULC classes simulated by the MCRF approach based on expert-interpreted data. Therefore, the proposed approach can improve classification accuracy of pre-classified maps as long as some expert-interpreted data are available. The main advantage of this approach is that it may comprehensively utilize a variety of available information seamlessly to LULC classification through expert interpretation and spatial cosimulation.

[1]  Peter M. Atkinson,et al.  Multiple-point geostatistical simulation for post-processing a remotely sensed land cover classification , 2013 .

[2]  Ming-Hseng Tseng,et al.  A genetic algorithm rule-based approach for land-cover classification , 2008 .

[3]  G. Matheron.  Principles of geostatistics , 1963 .

[4]  Chuanrong Zhang,et al.  A random-path markov chain algorithm for simulating categorical soil variables from random point samples , 2007 .

[5]  P. Gong,et al.  The use of structural information for improving land-cover classification accuracies at the rural-urban fringe. , 1990 .

[6]  Roussos Dimitrakopoulos,et al.  High-order Statistics of Spatial Random Fields: Exploring Spatial Cumulants for Modeling Complex Non-Gaussian and Non-linear Phenomena , 2009 .

[7]  Chuanrong Zhang,et al.  Restoration of the missing pixel information caused by contrails in multispectral remotely sensed imagery , 2014 .

[8]  T. Bayes.  An essay towards solving a problem in the doctrine of chances , 2003 .

[9]  Jon Atli Benediktsson,et al.  Neural Network Approaches Versus Statistical Methods in Classification of Multisource Remote Sensing Data , 1989, 12th Canadian Symposium on Remote Sensing Geoscience and Remote Sensing Symposium,.

[10]  Jessica Daecher,et al.  Gslib Geostatistical Software Library And Users Guide , 2016 .

[11]  Luis Ángel Ruiz Fernández,et al.  Definition of a comprehensive set of texture semivariogram features and their evaluation for object-oriented image classification , 2010, Comput. Geosci..

[12]  Amilcar Soares,et al.  Improving satellite images classification using remote and ground data integration by means of stochastic simulation , 2006 .

[13]  G. Fogg,et al.  Modeling Spatial Variability with One and Multidimensional Continuous-Lag Markov Chains , 1997 .

[14]  Weidong Li,et al.  Linear interpolation and joint model fitting of experimental transiograms for Markov chain simulation of categorical spatial variables , 2010, Int. J. Geogr. Inf. Sci..

[15]  Giles M. Foody,et al.  Status of land cover classification accuracy assessment , 2002 .

[16]  Antonio G. Chessa,et al.  A Markov Chain Model for Subsurface Characterization: Theory and Applications , 2006 .

[17]  Qihao Weng,et al.  A survey of image classification methods and techniques for improving classification performance , 2007 .

[18]  G. Matheron.  The intrinsic random functions and their applications , 1973, Advances in Applied Probability.

[19]  Xiuping Jia,et al.  Integrating remotely sensed data, GIS and expert knowledge to update object-based land use/land cover information , 2012 .

[20]  Freek D. van der Meer,et al.  Remote-sensing image analysis and geostatistics , 2012 .

[21]  Anders Rosholm,et al.  Statistical methods for segmentation and classification of images , 1997 .

[22]  M. Madden,et al.  Large area forest inventory using Landsat ETM+: A geostatistical approach , 2009 .

[23]  Anil K. Jain,et al.  A Markov random field model for classification of multisource satellite imagery , 1996, IEEE Trans. Geosci. Remote. Sens..

[24]  Junfeng Luo,et al.  Transition Probability Approach to Statistical Analysis of Spatial Qualitative Variables in Geology , 1996 .

[25]  H. Omre.  Bayesian kriging—Merging observations and qualified guesses in kriging , 1987 .

[26]  Mark E. Johnson,et al.  Multivariate Statistical Simulation , 1989, International Encyclopedia of Statistical Science.

[27]  John R. Jensen,et al.  Introductory Digital Image Processing: A Remote Sensing Perspective , 1986 .

[28]  G. Fogg,et al.  Transition probability-based indicator geostatistics , 1996 .

[29]  G. Christakos.  A Bayesian/maximum-entropy view to the spatial estimation problem , 1990 .

[30]  Alexandre Boucher,et al.  Sub-pixel Mapping of Coarse Satellite Remote Sensing Images with Stochastic Simulations from Training Images , 2009 .

[31]  Robert W. Ritzi,et al.  Behavior of indicator variograms and transition probabilities in relation to the variance in lengths of hydrofacies , 2000 .

[32]  Giles M. Foody,et al.  A relative evaluation of multiclass image classification by support vector machines , 2004, IEEE Transactions on Geoscience and Remote Sensing.

[33]  Graham E. Fogg,et al.  Multi-scale alluvial fan heterogeneity modeled with transition probability geostatistics in a sequence stratigraphic framework , 1999 .

[34]  Weidong Li,et al.  A Markov Chain Geostatistical Framework for Land-Cover Classification With Uncertainty Assessment Based on Expert-Interpreted Pixels From Remotely Sensed Imagery , 2011, IEEE Transactions on Geoscience and Remote Sensing.

[35]  P. Goovaerts.  Ordinary Cokriging Revisited , 1998 .

[36]  Pierre Goovaerts,et al.  Stochastic simulation of categorical variables using a classification algorithm and simulated annealing , 1996 .

[37]  S. Bruin.  Predicting the areal extent of land-cover types using classified imagery and geostatistics. , 2000 .

[38]  Sebastien Strebelle,et al.  Conditional Simulation of Complex Geological Structures Using Multiple-Point Statistics , 2002 .

[39]  Mark E. Johnson.  Multivariate Statistical Simulation: Johnson/Multivariate , 1987 .

[40]  Daniel F. Merriam,et al.  Geologic modeling and mapping , 1996 .

[41]  David B. Dunson,et al.  Bayesian Data Analysis , 2010 .

[42]  Weidong Li,et al.  Markov Chain Random Fields for Estimation of Categorical Variables , 2007 .

[43]  M. Ramsey,et al.  Monitoring urban land cover change: An expert system approach to land cover classification of semiarid to arid urban centers , 2001 .

[44]  Yong Ge,et al.  Multiple-point simulation-based method for extraction of objects with spatial structure from remotely sensed imagery , 2011 .

[45]  Michel Dekking,et al.  A Markov Chain Model for Subsurface Characterization: Theory and Applications , 2001 .

[46]  P. Swain,et al.  Neural Network Approaches Versus Statistical Methods In Classification Of Multisource Remote Sensing Data , 1990 .

[47]  Weidong Li,et al.  Transiograms for Characterizing Spatial Variability of Soil Classes , 2007 .

[48]  D. K. Pickard,et al.  Unilateral Markov fields , 1980, Advances in Applied Probability.

[49]  W. Schwarzacher,et al.  The use of Markov chains in the study of sedimentary cycles , 1969 .

[50]  Dipak K. Dey,et al.  Updating Categorical Soil Maps Using Limited Survey Data by Bayesian Markov Chain Cosimulation , 2013, TheScientificWorldJournal.