Building footprint maps are vital to many remote sensing (RS) applications, such as 3-D building modeling, urban planning, and disaster management. Due to the complexity of buildings, the accurate ...and reliable generation of the building footprint from RS imagery is still a challenging task. In this article, an end-to-end building footprint generation approach that integrates convolution neural network (CNN) and graph model is proposed. CNN serves as the feature extractor, while the graph model can take spatial correlation into consideration. Moreover, we propose to implement the feature pairwise conditional random field (FPCRF) as a graph model to preserve sharp boundaries and fine-grained segmentation. Experiments are conducted on four different data sets: 1) Planetscope satellite imagery of the cities of Munich, Paris, Rome, and Zurich; 2) ISPRS Benchmark data from the city of Potsdam; 3) Dstl Kaggle data set; and 4) Inria Aerial Image Labeling data of Austin, Chicago, Kitsap County, Western Tyrol, and Vienna. It is found that the proposed end-to-end building footprint generation framework with the FPCRF as the graph model can further improve the accuracy of building footprint generation by using only CNN, which is the current state of the art.
Thrombosis has become a global disease threatening human health. The left atrial appendage (LAA) is a major source of thrombosis in patients with atrial fibrillation (AF). Positive correlation exists ...between LAA volume and AF risk. LAA morphology has been suggested to influence thromboembolic risk in AF patients and to help predict thromboembolic events in low-risk patient groups. Automatic segmentation of LAA can greatly help physicians diagnose AF. In consideration of the large anatomical variations of the LAA, we proposed a robust method for automatic LAA segmentation on computed tomographic angiography (CTA) data using fully convolutional neural networks with three-dimensional (3-D) conditional random fields (CRFs). After manual localization of ROI of LAA, we adopted the FCN in natural image segmentation and transferred their learned models by fine-tuning the networks to segment each 2-D LAA slice. Subsequently, we used a modified dense 3-D CRF that accounts for the 3-D spatial information and larger contextual information to refine the segmentations of all slices. Our method was evaluated on 150 sets of CTA data using five-fold cross validation. Compared with manual annotation, we obtained a mean dice overlap of 94.76% and a mean volume overlap of 91.10% with a computation time of less than 40 s per volume. Experimental results demonstrated the robustness of our method in dealing with large anatomical variations and computational efficiency for adoption in a daily clinical routine.
Recent advances on adaptive steganography show that the performance of image steganographic communication can be improved by incorporating the non-additive models that capture the dependencies among ...adjacent pixels. In this paper, a Gaussian Markov Random Field model (GMRF) with four-element cross neighborhood is proposed to characterize the interactions among local elements of cover images, and the problem of secure image steganography is formulated as the one of minimization of KL-divergence in terms of a series of low-dimensional clique structures associated with GMRF by taking advantages of the conditional independence of GMRF. The adoption of the proposed GMRF tessellates the cover image into two disjoint subimages, and an alternating iterative optimization scheme is developed to effectively embed the given payload while minimizing the total KL-divergence between cover and stego, i.e., the statistical detectability. Experimental results demonstrate that the proposed GMRF outperforms the prior arts of model based schemes, e.g., MiPOD, and rivals the state-of-the-art HiLL for practical steganography, where the selection channel knowledges are unavailable to steganalyzers.
A general form of the covariance matrix function is derived in this paper for a vector random field that is isotropic and mean square continuous on a compact connected two-point homogeneous space and ...stationary on a temporal domain. A series representation is presented for such a vector random field which involves Jacobi polynomials and the distance defined on the compact two-point homogeneous space.
Accurate and precise characterization of the subsurface stratigraphic configuration (geological model) at a given site is crucial to geotechnical engineering work. The uncertainty in the derived ...stratigraphic configuration can be significant, due to the strata's complexity and inherent spatial variability coupled with the limited availability of borehole data. The characterization and reduction of this uncertainty should be part of any site characterization project. This paper presents a method for characterization of the subsurface stratigraphic configuration with limited borehole data. Within the framework of the proposed method, the spatial correlation between the existence of a stratum in one subsurface zone and that in the other subsurface zone (or the spatial correlation of the existence of the stratum) is captured by an autocorrelation function determined with the maximum likelihood principle. The initial stratigraphic configurations are first sampled with the conditional random field theory. Next, the maximum-a-posteriori (MAP) estimates of the initial stratigraphic configurations are derived using Markov Chain Monte Carlo (MCMC) and taken as the final stratigraphic realizations. The effectiveness of the proposed method and its advantages over the existing stratigraphic characterization methods are demonstrated through a series of comparative analyses. The versatility of the new approach in modeling the 3-D stratigraphic configuration is further revealed through a case study of a site in Western Australia. This paper adds to the literature on stratigraphic uncertainty characterization and provides a basis for a risk-based geotechnical assessment that considers geological and geotechnical uncertainties.
•A probabilistic method for characterizing stratigraphic configuration is proposed.•The spatial correlation of strata is estimated with maximum likelihood principle.•Conditional random field theory and MCMC method are included in this method.•Versatility of this method in modeling 3-D stratigraphic configurations is studied.
Automatic segmentation of lung lesions from COVID-19 computed tomography (CT) images can help to establish a quantitative model for diagnosis and treatment. For this reason, this article provides a ...new segmentation method to meet the needs of CT images processing under COVID-19 epidemic. The main steps are as follows: First, the proposed region of interest extraction implements patch mechanism strategy to satisfy the applicability of 3-D network and remove irrelevant background. Second, 3-D network is established to extract spatial features, where 3-D attention model promotes network to enhance target area. Then, to improve the convergence of network, a combination loss function is introduced to lead gradient optimization and training direction. Finally, data augmentation and conditional random field are applied to realize data resampling and binary segmentation. This method was assessed with some comparative experiment. By comparison, the proposed method reached the highest performance. Therefore, it has potential clinical applications.
In this article, we tackle the problem of depth estimation from single monocular images. Compared with depth estimation using multiple images such as stereo depth perception, depth from monocular ...images is much more challenging. Prior work typically focuses on exploiting geometric priors or additional sources of information, most using hand-crafted features. Recently, there is mounting evidence that features from deep convolutional neural networks (CNN) set new records for various vision applications. On the other hand, considering the continuous characteristic of the depth values, depth estimation can be naturally formulated as a continuous conditional random field (CRF) learning problem. Therefore, here we present a deep convolutional neural field model for estimating depths from single monocular images, aiming to jointly explore the capacity of deep CNN and continuous CRF. In particular, we propose a deep structured learning scheme which learns the unary and pairwise potentials of continuous CRF in a unified deep CNN framework. We then further propose an equally effective model based on fully convolutional networks and a novel superpixel pooling method, which is about 10 times faster, to speedup the patch-wise convolutions in the deep model. With this more efficient model, we are able to design deeper networks to pursue better performance. Our proposed method can be used for depth estimation of general scenes with no geometric priors nor any extra information injected. In our case, the integral of the partition function can be calculated in a closed form such that we can exactly solve the log-likelihood maximization. Moreover, solving the inference problem for predicting depths of a test image is highly efficient as closed-form solutions exist. Experiments on both indoor and outdoor scene datasets demonstrate that the proposed method outperforms state-of-the-art depth estimation approaches.
The present paper presents some results that allow us to perform with High Relative Accuracy linear algebra operations with correlation and covariance matrices of Gaussian Markov Random Fields over ...graphs of paths. Some numerical experiments are carried out showing the computational benefits of this approach.
This paper proposes an effective multi-view learning approach to foreground detection for traffic surveillance applications. This approach involves three main steps. First, a reference background ...image is generated via temporal median filtering, and multiple heterogeneous features (including brightness variation, chromaticity variation, and texture variation, each of which represents a unique view) are extracted from the video sequence. Then, a multi-view learning strategy is devised to online estimate the conditional probability densities for both the foreground and the background. The probability densities of three features are approximately conditionally independent and are estimated with kernel density estimation. Pixel soft labeling is conducted by using Bayes rule, and the pixelwise foreground posteriors are computed. Finally, a Markov random field is constructed to incorporate the spatiotemporal context into the foreground/background decision model. The belief propagation algorithm is used to label each pixel of the current frame. Experimental results verify that the proposed approach is effective to detect foreground objects from challenging traffic environments and outperforms some state-of-the-art methods.
•We propose a novel retinal vessel segmentation method of fundus images.•The method is combined with CNN and fully CRFs.•We develop a multiscale CNN with an improved cross-entropy loss function.•We ...evaluate the effectiveness of the proposed method on DRIVE and STARE datasets.•The results of our method are competitive in the sensitivity while ensuring accuracy.
Retinal vessel analysis of fundus images is an indispensable method for the screening and diagnosis of related diseases. In this paper, we propose a novel retinal vessel segmentation method of the fundus images based on convolutional neural network (CNN) and fully connected conditional random fields (CRFs). The segmentation process is mainly divided into two steps. Firstly, a multiscale CNN architecture with an improved cross-entropy loss function is proposed to produce the probability map from image to image. We construct the multiscale network by combining the feature map of each middle layer to learn more detail information of the retinal vessels. Meanwhile, our proposed cross-entropy loss function ignores the slightest loss of relatively easy samples in order to take more attention to learn the hard examples. Secondly, CRFs is applied to get the final binary segmentation result which makes use of more spatial context information by taking into account the interactions among all of the pixels in the fundus images. The effectiveness of the proposed method has been evaluated on two public datasets, i.g., DRIVE and STARE with comparisons against eleven state-of-the-art approaches including five deep learning based methods. Results show that our method allows for detection of more tiny blood vessels and more precise locating of the edges.
PDF
