Markov Random Field Approach Research Articles

Interactive Image Segmentation using Improved Adaptive Markov Random Field Approach

Background/Objectives: To interactively split an object of interest from the remaining image with better smoothing by introducing improved adaptive Markov Random field resolving the problem of much noise. Methods/Statistical analysis: By utilizing a Dirichlet process multiple - view learning scheme, the unlabelled pixel labels are calculated by using the seed pixels. It is used for supporting the multiple-view learning in order to incorporate the constraints of both appearance and boundary constraint, and the Dirichlet process mixture-based nonlinear classification for concurrently modelling the image features and distinguishing the differences between the classes of object and background. The MRF field is utilized for providing the smoothness in segment labels. Findings: In Markov Random Field (MRF) based scheme, only the pixel and the surrounding pixels relationships are considered. The microscopic image processing result along with low noise is bad. To solve this problem, the adaptive MRF method based on region; it exploits Graph Cuts for inference. The different type of images produces the different pre-segmentation results. The connection degree between current and its linked regional blocks is denoted by connection parameter. If the connection parameter value is high, large connection degree between regional block and neighbouring regional blocks is defined. Otherwise lower connection degree between regional block and neighbourhood regional blocks. The adaptive MRF smoothing is more accurate and efficient segmentation result than the traditional MRF based Smoothing method. However, the appropriate parameter selection is a difficult task for practical image segmentation which can be solved by introducing an improved adaptive MRF model by using a modified graph cutter. Thus, the image segmentation is achieved based on modified graph-cut model using a novel energy function without the regularizing parameter. Improvements/Applications: The improved adaptive Markov Random Field approach interactively segment images with better smoothness than most of the current approaches.

SAR Image segmentation based on convolutional-wavelet neural network and markov random field

Synthetic aperture radar (SAR) imaging system is usually an observation of the earths' surface. It means that rich structures exist in SAR images. Convolutional neural network (CNN) is good at learning features from raw data automatically, especially the structural features. Inspired by these, we propose a novel SAR image segmentation method based on convolutional-wavelet neural networks (CWNN) and Markov Random Field (MRF). In this approach, a wavelet constrained pooling layer is designed to replace the conventional pooling in CNN. The new architecture can suppress the noise and is better at keeping the structures of the learned features, which are crucial to the segmentation tasks. CWNN produces the segmentation map by patch-by-patch scanning. The segmentation result of CWNN will be used with two labeling strategies (i.e., a superpixel approach and a MRF approach) to produce the final segmentation map. The superpixel approach is used to enforce the smooth nature on the local region. On the other hand, the MRF approach is used to preserve the edges and the details of the SAR image. Specifically, two segmentation maps will be produced by applying the superpixel and MRF approaches. The first segmentation map is obtained by combining the segmentation map of CWNN and the superpixel approach, and the second segmentation map is obtained by applying the MRF approach on the original SAR image. Afterwards, these two segmentation maps are fused by using the sketch map of the SAR image to produce the final segmentation map. Experiments on the texture images demonstrate that the CWNN is effective for the segmentation tasks. Moreover, the experiments on the real SAR images show that our approach obtains the regions with labeling consistency and preserves the edges and details at the same time.

Toward a Sparse Bayesian Markov Random Field Approach to Hyperspectral Unmixing and Classification.

Recent work has shown that existing powerful Bayesian hyperspectral unmixing algorithms can be significantly improved by incorporating the inherent local spatial correlations between pixel class labels via the use of Markov random fields. We here propose a new Bayesian approach to joint hyperspectral unmixing and image classification such that the previous assumption of stochastic abundance vectors is relaxed to a formulation whereby a common abundance vector is assumed for pixels in each class. This allows us to avoid stochastic reparameterizations and, instead, we propose a symmetric Dirichlet distributionmodel with adjustable parameters for the common abundance vector of each class. Inference over the proposed model is achieved via a hybrid Gibbs sampler, and in particular, simulated annealing is introduced for the label estimation in order to avoid the local-trap problem. Experiments on a synthetic image and a popular, publicly available real data set indicate the proposed model is faster than and outperforms the existing approach quantitatively and qualitatively. Moreover, for appropriate choices of the Dirichlet parameter, it is shown that the proposed approach has the capability to induce sparsity in the inferred abundance vectors. It is demonstrated that this offers increased robustness in cases where the preprocessing endmember extraction algorithms overestimate the number of active endmembers present in a given scene.

Facial Sketch Synthesis Using 2D Direct Combined Model-Based Face-Specific Markov Network.

A facial sketch synthesis system is proposed, featuring a 2D direct combined model (2DDCM)-based face-specific Markov network. In contrast to the existing facial sketch synthesis systems, the proposed scheme aims to synthesize sketches, which reproduce the unique drawing style of a particular artist, where this drawing style is learned from a data set consisting of a large number of image/sketch pairwise training samples. The synthesis system comprises three modules, namely, a global module, a local module, and an enhancement module. The global module applies a 2DDCM approach to synthesize the global facial geometry and texture of the input image. The detailed texture is then added to the synthesized sketch in a local patch-based manner using a parametric 2DDCM model and a non-parametric Markov random field (MRF) network. Notably, the MRF approach gives the synthesized results an appearance more consistent with the drawing style of the training samples, while the 2DDCM approach enables the synthesis of outcomes with a more derivative style. As a result, the similarity between the synthesized sketches and the input images is greatly improved. Finally, a post-processing operation is performed to enhance the shadowed regions of the synthesized image by adding strong lines or curves to emphasize the lighting conditions. The experimental results confirm that the synthesized facial images are in good qualitative and quantitative agreement with the input images as well as the ground-truth sketches provided by the same artist. The representing power of the proposed framework is demonstrated by synthesizing facial sketches from input images with a wide variety of facial poses, lighting conditions, and races even when such images are not included in the training data set. Moreover, the practical applicability of the proposed framework is demonstrated by means of automatic facial recognition tests.

A Markov random field approach for microstructure synthesis

We test the notion that many microstructures have an underlying stationary probability distribution. The stationary probability distribution is ubiquitous: we know that different windows taken from a polycrystalline microstructure are generally ‘statistically similar’. To enable computation of such a probability distribution, microstructures are represented in the form of undirected probabilistic graphs called Markov Random Fields (MRFs). In the model, pixels take up integer or vector states and interact with multiple neighbors over a window. Using this lattice structure, algorithms are developed to sample the conditional probability density for the state of each pixel given the known states of its neighboring pixels. The sampling is performed using reference experimental images. 2D microstructures are artificially synthesized using the sampled probabilities. Statistical features such as grain size distribution and autocorrelation functions closely match with those of the experimental images. The mechanical properties of the synthesized microstructures were computed using the finite element method and were also found to match the experimental values.

A Hybrid Particle Swarm Optimization with Affine Transformation Approach for Cloud Free Multi-Temporal Image Registration

An image registration is the major part of the image categorization and cluster formation in multi temporal image processing. The images are affected by the different factors such as cloud shadow, water level, building shadows etc. In this paper, an enhanced registration process and the cloud removal technique is proposed for image enhancement. The Daemons, Combined Registration and Segmentation (CRS) approach, Markov Random Field (MRF) approach and Mutual Information (MI) based approaches results in more computational complexity, minimum edge preservation measure (QAB/F) and Mutual Information in image registration. In order to maximize the quality of edge preservation measure and MI with minimum computational time, this paper proposes Particle Swarm Optimization (PSO) based affine transformation technique. The proposed techniques measure and compare the computation time against the number of pixels of an image with the existing methods of CRS and MRF for the number of images. The comparative analysis of QAB/F and MI with the traditional methods of Clock Point –Least Square (CP-LS) and the Multi-Focus Image Fusion (MFIF) and Discrete Wavelet Transform (DWT) is presented to confirm the effective performance. The simulation results of the proposed transformation for registration process confirms the effective image registration in the multi-temporal image processing.

3D MICROWAVE TOMOGRAPHY WITH HUBER REGULARIZATION APPLIED TO REALISTIC NUMERICAL BREAST PHANTOMS

Quantitative active microwave imaging for breast cancer screening and therapy monitoring applications requires adequate reconstruction algorithms, in particular with regard to the nonlinearity and ill-posedness of the inverse problem. We employ a fully vectorial three-dimensional nonlinear inversion algorithm for reconstructing complex permittivity profiles from multi-view single-frequency scattered field data, which is based on a Gauss-Newton optimization of a regularized cost function. We tested it before with various types of regularizing functions for piecewise-constant objects from Institut Fresnel and with a quadratic smoothing function for a realistic numerical breast phantom. In the present paper we adopt a cost function that includes a Huber function in its regularization term, relying on a Markov Random Field approach. The Huber function favors spatial smoothing within homogeneous regions while preserving discontinuities between contrasted tissues. We illustrate the technique with 3D reconstructions from synthetic data at 2GHz for realistic numerical breast phantoms from the University of Wisconsin-Madison UWCEM online repository: we compare Huber regularization with a multiplicative smoothing regularization and show reconstructions for various positions of a tumor, for multiple tumors and for different tumor sizes, from a sparse and from a denser data configuration.

Semantic volume segmentation with iterative context integration for bio-medical image stacks

Automatic recognition of biological structures like membranes or synapses is important to analyze organic processes and to understand their functional behavior. To achieve this, volumetric images taken by electron microscopy or computer tomography have to be segmented into meaningful semantic regions. We are extending iterative context forests which were developed for 2D image data to image stack segmentation. In particular, our method is able to learn high-order dependencies and import contextual information, which often can not be learned by conventional Markov random field approaches usually used for this task. Our method is tested on very different and challenging medical and biological segmentation tasks.

A Markov random field approach to group-wise registration/mosaicing with application to ultrasound.

In this paper we present a group-wise non-rigid registration/mosaicing algorithm based on block-matching, which is developed within a probabilistic framework. The discrete form of its energy functional is linked to a Markov Random Field (MRF) containing double and triple cliques, which can be effectively optimized using modern MRF optimization algorithms popular in computer vision. Also, the registration problem is simplified by introducing a mosaicing function which partitions the composite volume into regions filled with data from unique, partially overlapping source volumes. Ultrasound confidence maps are incorporated into the registration framework in order to give accurate results in the presence of image artifacts. The algorithm is initially tested on simulated images where shadows have been generated. Also, validation results for the group-wise registration algorithm using real ultrasound data from an abdominal phantom are presented. Finally, composite obstetrics image volumes are constructed using clinical scans of pregnant subjects, where fetal movement makes registration/mosaicing especially difficult. In addition, results are presented suggesting that a fusion approach to MRF registration can produce accurate displacement fields much faster than standard approaches.

Statistically-driven 3D fiber reconstruction and denoising from multi-slice cardiac DTI using a Markov random field model.

The construction of subject-specific dense and realistic 3D meshes of the myocardial fibers is an important pre-requisite for the simulation of cardiac electrophysiology and mechanics. Current diffusion tensor imaging (DTI) techniques, however, provide only a sparse sampling of the 3D cardiac anatomy based on a limited number of 2D image slices. Moreover, heart motion affects the diffusion measurements, thus resulting in a significant amount of noisy fibers. This paper presents a Markov random field (MRF) approach for dense reconstruction of 3D cardiac fiber orientations from sparse DTI 2D slices. In the proposed MRF model, statistical constraints are used to relate the missing and the known fibers, while a consistency term is encoded to ensure that the obtained 3D meshes are locally continuous. The validation of the method using both synthetic and real DTI datasets demonstrates robust fiber reconstruction and denoising, as well as physiologically meaningful estimations of cardiac electrical activation.

An Automatic <inline-formula> <tex-math notation="TeX">${\cal U}$</tex-math></inline-formula>-Distribution and Markov Random Field Segmentation Algorithm for PolSAR Images

We have recently presented a novel unsupervised, non-Gaussian, and contextual clustering algorithm for segmentation of polarimetric synthetic aperture radar (PolSAR) images. This represents one of the most advanced PolSAR unsupervised statistical segmentation algorithms and uses the doubly flexible two-parameter U-distribution model for the PolSAR statistics and includes a Markov random field (MRF) approach for contextual smoothing. A goodness-of-fit testing stage adds a statistically rigorous approach to determine the significant number of classes. The fully automatic algorithm was demonstrated with good results for both simulated and real data sets. This paper discusses a rethinking of the overall strategy and leads to some simplifications. The primary issue was that the MRF optimization depends on the number of classes and did not behave well under the split-and-merge environment. We explain the reasons behind a separation of the cluster evaluation from the contextual smoothing and a modified rationale for the adaptive number of classes. Both aspects have simplified the overall algorithm while maintaining good visual results.

An Unsupervised Change Detection in Satellite IMAGES Using MRFFCM Clustering

This paper presents a new approach for change detection in synthetic aperture radar images by incorporating Markov random field (MRF) within the framework of FCM. The objective is to partition the difference image which is generated from multitemporal satellite images into changed and unchanged regions. The difference image is generated from log ratio and mean ratio images by image fusion technique. The quality of difference image depends on image fusion technique. In the present work; we have proposed an image fusion method based on stationary wavelet transform. To process the difference image is to discriminate changed regions from unchanged regions using fuzzy clustering algorithms. The analysis of the DI is done using Markov random field (MRF) approach that exploits the interpixel class dependency in the spatial domain to improve the accuracy of the final change-detection areas. The experimental results on real synthetic aperture radar images demonstrate that change detection results obtained by the MRFFCM exhibits less error than previous approaches. The goodness of the proposed fusion algorithm by well-known image fusion measures and the percentage correct classifications are calculated and verified. Keywords-image change detection, MMRM, LRM, MRFFCM, SAR, fuzzy clustering

SPATIAL-TEMPORAL CONDITIONAL RANDOM FIELDS CROP CLASSIFICATION FROM TERRASAR-X IMAGES

Abstract. The rapid increase in population in the world has propelled pressure on arable land. Consequently, the food basket has continuously declined while global demand for food has grown twofold. There is need to monitor and update agriculture land-cover to support food security measures. This study develops a spatial-temporal approach using conditional random fields (CRF) to classify co-registered images acquired in two epochs. We adopt random forest (RF) as CRF association potential and introduce a temporal potential for mutual crop phenology information exchange between spatially corresponding sites in two epochs. An important component of temporal potential is a transitional matrix that bears intra- and inter-class changes between considered epochs. Conventionally, one matrix has been used in the entire image thereby enforcing stationary transition probabilities in all sites. We introduce a site dependent transition matrix to incorporate phenology information from images. In our study, images are acquired within a vegetation season, thus perceived spectral changes are due to crop phenology. To exploit this phenomena, we develop a novel approach to determine site-wise transition matrix using conditional probabilities computed from two corresponding temporal sites. Conditional probability determines transitions between classes in different epochs and thus we used it to propagate crop phenology information. Classification results show that our approach improved crop discrimination in all epochs compared to state-of-the-art mono-temporal approaches (RF and CRF monotemporal) and existing multi-temporal markov random fields approach by Liu et al. (2008).

Stochastic Path Model of Polaroid Polarizer for Bell's Theorem and Triphoton Experiments

Depending on the outcome of the triphoton experiment now underway, it is possible that the new local realistic Markov Random Field (MRF) models will be the only models now available to correctly predict both that experiment and Bell's theorem experiments. The MRF models represent the experiments as graphs of discrete events over space-time. This paper extends the MRF approach to continuous time, by defining a new class of realistic model, the stochastic path model, and showing how it can be applied to ideal polaroid type polarizers in such experiments. The final section discusses possibilities for future research, ranging from uses in other experiments or novel quantum communication systems, to extensions involving stochastic paths in the space of functions over continuous space. As part of this, it derives a new Boltzmann-like density operator over Fock space, which predicts the emergent statistical equilibria of nonlinear Hamiltonian field theories, based on our previous work of extending the Glauber–Sudarshan P mapping from the case of classical systems described by a complex state variable α to the case of classical continuous fields. This extension may explain the stochastic aspects of quantum theory as the emergent outcome of nonlinear PDE in a time-symmetric universe.

Texture Analysis Based on Micro Primitive Descriptor (MPD)

Texture classification is an important application in all the fields of image processing and computer vision. This paper proposes a simple and powerful feature set for texture classification, namely micro primitive descriptor (MPD). The MPD is derived from the 2×2 grid of a motif transformed image. The original image is divided into 2×2 pixel grids. Each 2×2 grid is replaced by a motif shape that minimizes the local ascent while traversing the 2×2 grid forming a motif transformed image. The proposed feature set extracts textural information of an image with a more detailed respect of texture characteristics. The results demonstrate that it is much more efficient and effective than representative feature descriptors, such as Random Threshold Vector Technique (RTV) features and Wavelet Transforms Based on Gaussian Markov Random Field (WTBGMF) approach for texture classification. Index Terms—Micro primitive descriptor, motif transformed image, texture classification.

Improved adaptive Markov random field based super-resolution mapping for mangrove tree identification

Abstract. Traditionally, forest tree crowns are extracted using airborne or spaceborne hyper-/multi-spectral remotely sensed images or pansharpened images. However, these medium/low spatial resolution images suffer from the mixed pixel problem, and the cost to collect very high resolution image collection is high. Moreover, existing feature extraction techniques cannot extract local patterns from medium/low resolution images. Therefore, super-resolution mapping (SRM) techniques, which generate land-cover maps with finer spatial resolution than the original remotely sensed image, can be beneficial for the extraction of forest trees. The SRM methods can improve the quality of information extraction by combining spectral information and spatial context into image classification problems. In this paper we have improved an adaptive Markov random field approach for super-resolution mapping (MRF-SRM) based on spatially adaptive MRF-SPM to overcome the limitation of equal covariance matrices assumption for all classes. We applied the developed method for mangrove tree identification from multispectral image recorded by QuickBird satellite, where we generated a super-resolution map with the panchromatic image spatial resolution of 0.6 m. Moreover, the performance of the proposed technique is evaluated by employing the simulated image with different covariance matrices for each class. Our experimental results have demonstrated that the new adaptive MRF-SRM method has increased the overall accuracy by 5.1% and the termination conditions of this method were satisfied three times faster when compared to the state-of-the-art methods.

Fast pose invariant face recognition using super coupled multiresolution Markov Random Fields on a GPU

We discuss the problem of pose invariant face recognition using a Markov Random Field (MRF) model. MRF image to image matching has been shown to be very promising in earlier studies (Arashloo and Kittler, 2011) [4]. Its demanding computational complexity has been addressed in Arashloo et al. (2011) [6] by means of multiresolution MRFs linked by the super coupling transform advocated by Petrou et al. (1998) [37, 11]. In this paper, we benefit from the daisy descriptor for face image representation in image matching. Most importantly, we design an innovative GPU implementation of the proposed multiresolution MRF matching process. The significant speed up achieved (factor of 25) has multiple benefits: It makes the MRF approach a practical proposition. It facilitates extensive empirical optimisation and evaluation studies. The latter conducted on benchmarking databases, including the challenging labelled faces in the wild (LFW) database show the outstanding potential of the proposed method, which consistently achieves state-of-the-art performance in standard benchmarking tests. The experimental studies also show that the super coupled multiresolution MRFs deliver a computational speed up by a factor of 5 over and above the speed up achieved using the GPU implementation.

Spatial-Attraction-Based Markov Random Field Approach for Classification of High Spatial Resolution Multispectral Imagery

This letter presents a novel spatial-attraction-based Markov random field (MRF) (SAMRF) approach for high spatial resolution multispectral imagery (HSRMI) classification. First, the initial class label and class membership for each pixel are obtained by applying the maximum likelihood classifier (MLC) classification for the HSRMI. Second, to reduce the oversmooth classification in the traditional MRF, an adaptive weight MRF model is introduced by integrating the spatial attraction model into the traditional MRF. Finally, the initial classification map, generated in the first step, will be refined though the SAMRF regularization. Two different experiments were performed to evaluate the performance of the SAMRF, in comparison with standard MLC and MRF. Experimental results indicate that the SAMRF method achieved the highest accuracy, hence, providing an effective spectral-spatial classification method for the HSRMI.

Resolving the Antarctic contribution to sea-level rise: a hierarchical modelling framework.

Determining the Antarctic contribution to sea-level rise from observational data is a complex problem. The number of physical processes involved (such as ice dynamics and surface climate) exceeds the number of observables, some of which have very poor spatial definition. This has led, in general, to solutions that utilise strong prior assumptions or physically based deterministic models to simplify the problem. Here, we present a new approach for estimating the Antarctic contribution, which only incorporates descriptive aspects of the physically based models in the analysis and in a statistical manner. By combining physical insights with modern spatial statistical modelling techniques, we are able to provide probability distributions on all processes deemed to play a role in both the observed data and the contribution to sea-level rise. Specifically, we use stochastic partial differential equations and their relation to geostatistical fields to capture our physical understanding and employ a Gaussian Markov random field approach for efficient computation. The method, an instantiation of Bayesian hierarchical modelling, naturally incorporates uncertainty in order to reveal credible intervals on all estimated quantities. The estimated sea-level rise contribution using this approach corroborates those found using a statistically independent method. © 2013 The Authors. Environmetrics Published by John Wiley & Sons, Ltd.

CONTEXTUAL IMAGE CLASSIFICATION APPROACH FOR MONITORING OF AGRICULTURAL LAND COVER BY SUPPORT VECTOR MACHINES AND MARKOV RANDOM FIELDS

Abstract. The main idea of this paper is to integrate the non-contextual support vector machines (SVM) classifiers with Markov random fields (MRF) approach to develop a contextual framework for monitoring of agricultural land cover. To this end, the SVM and MRF approaches were integrated to exploit both spectral and spatial contextual information in the image for more accurate classification of remote sensing data from an agricultural region in Biddinghuizen, the Netherlands. Comparative analysis of this study clearly demonstrated that the proposed contextual method based on SVM-MRF models generates a higher average accuracy, overall accuracy and Kappa coefficient compared with non-contextual SVM method. Since the spatial information is considered in the proposed method, this study indicates that a neater, more homogonous and speckle-free results could be generated by the SVM-MRF approach.