As a common infrastructure, roads, subject to heavy traffic in modern cities, are progressively becoming dangerous since they are under the threat of subsurface disasters caused by geological structure changes, groundwater overutilization, etc. Benefiting from the superiority in nondestructivity as well as efficiency, ground-penetrating radar (GPR) has been widely applied in the detection of underground disasters; however, manual interpretation of GPR images is labor-intensive and the intelligentization of GPR is impeded by the lack of labeled images. Considering that the texture pattern of echo waves in real b-scan images is the key clue for engineers to classify underground disasters, we proposed a semantic synthesis network with a high-frequency structure, i.e., HiFE-GAN, to augment GPR b-scan images for training detection networks, by which one can customize the position, scale, and size of underground disasters in the simulated b-scan images. The provided experiments show that our network can generate b-scan images more similar to the real ones compared with several current generation networks; meanwhile, mixing the simulated b-scan images synthesized by HiFE-GAN with the real ones to train detection networks can dramatically improve the detection performance for underground disasters.

Being one of infrastructures for modern cities, urban roads suffer from potential subsurface disasters, which results in unexpected vital loss of economy and life. Thanks to the superiority of nondestructive detection with high efficiency, ground penetrating radar (GPR) has been widely applied to underground disaster detection; however, fenced by lack of labelled GPR data, automatic detection methods, especially the ones based on deep neural networks, have to be trained by synthetic GPR images and few real ones, which impedes the further application of deep neural networks in underground disaster detection. We proposed a network based on generative adversarial network with multiscale discrimination strategy to generate GPR b-scan images from the synthetic images, i.e., the forward GPR b-scan images generated by gprMax. Because sharing the same physical laws with real b-scan images and carrying label information of subsurface disaster, the simulated images could be utilized to augment training dataset for detection networks. The associated experiments show that the simulated images by our network are very similar to the real GPR b-scan images in appearance; meanwhile, the detection networks trained on the data set mixing the b-scan images simulated by our network and real ones could achieve better performance. Using our network as an augmentation method for GPR b-scan images contributes to the extensive application of deep neural networks in intelligent processing of GPR data.

There are two problems in SAR image change detection when using difference images (DIs), i.e., 1) the subtraction operation results in serious loss of semantic information in DIs; and 2) the boundary of DI is uncertain. We propose a change detection method based on Neural-CRF structure. Firstly, Transformer-UNet (TR-UNet) is designed to provide the unary potential for CRF. The TR-Attention module improves the semantic expression ability of UNet by introducing the multi-head attention mechanism of TR. Secondly, a cascade CRF as Recurrent Neural Network, named as C-CRF-RNN, is proposed to update the unary potential and pairwise potential simultaneously. This improves the ability of CRF-RNN to refine pixel-level label prediction. Experiments show that the proposed method consistently outperforms the state-of-the-art methods on two benchmarks including berne data and ottawa data.

We proposed a contour co-tracking method for co-segmentation of image pairs based on active contour model. Our method comprehensively re-models objects and backgrounds signified by level set functions, and leverages Hellinger distance to measure the similarity between image regions encoded by probability distributions. The main contribution are as follows. 1) The new energy functional, combining a rewarding and a penalty term, relaxes the assumptions of co-segmentation methods. 2) Hellinger distance, fulfilling the triangle inequality, ensures a coherence measurement between probability distributions in metric space, and contributes to finding a unique solution to the energy functional. The proposed contour co-tracking method was carefully verified against five representative methods on four popular datasets, i.e., the images pair dataset (105 pairs), MSRC dataset (30 pairs), iCoseg dataset (66 pairs) and Coseg-rep dataset (25 pairs). The comparison experiments suggest that our method achieves the competitive and even better performance compared to the state-of-the-art co-segmentation methods.

This paper presents a hybrid level set method for object segmentation. The method deconstructs segmentation task into two procedures, i.e., shape transformation and curve evolution, which are alternately optimized until convergence. In this framework, only one shape prior encoded by shape context is utilized to estimate a transformation allowing the curve to have the same semantic expression as shape prior, and curve evolution is driven by an energy functional with topology-preserving and kernelized terms. In such a way, the proposed method is featured by the following advantages: 1) hybrid paradigm makes the level set framework possess the ability of incorporating other shape-related techniques about shape descriptor and distance; 2) shape context endows one single prior with semanticity, and hence leads to the competitive performance compared to the ones with multiple shape priors; and 3) additionally, combining topology-preserving and kernelization mechanisms together contributes to realizing a more reasonable segmentation on textured and noisy images. As far as we know, we propose a hybrid level set framework and utilize shape context to guide curve evolution for the first time. Our method is evaluated with synthetic, healthcare, and natural images, as a result, it shows competitive and even better performance compared to the counterparts.

In the field of connectomics, neuroscientists seek to identify cortical connectivity comprehensively. Neuronal boundary detection from the Electron Microscopy (EM) images is often done to assist the automatic reconstruction of neuronal circuit. But the segmentation of EM images is a challenging problem, as it requires the detector to be able to detect both filament-like thin and blob-like thick membrane, while suppressing the ambiguous intracellular structure. In this paper, we propose multi-stage multi-recursiveinput fully convolutional networks to address this problem. The multiple recursive inputs for one stage, i.e., the multiple side outputs with different receptive field sizes learned from the lower stage, provide multi-scale contextual boundary information for the consecutive learning. This design is biologically-plausible, as it likes a human visual system to compare different possible segmentation solutions to address the ambiguous boundary issue. Our multi-stage networks are trained end-to-end. It achieves promising results on two public available EM segmentation datasets, the mouse piriform cortex dataset and the ISBI 2012 EM dataset.

In this paper, inclusion degree of fuzzy sets is introduced to image segmentation. The image segmentation problem is novelly modeled as the minimization of the overlapping rates between the inside and outside regions, subject to a constraint on the total length of the region boundaries. Considering the similar properties of fuzzy sets and statistical image domain, we use fuzzy membership functions to represent the inside and outside regions and utilize nonparametric density estimates to estimate them. Then, the inclusion degree of fuzzy sets is adopted to formulate the overlapping rates between the inside and outside regions. We solve the inclusion-degree-based optimization problem by deriving the associated gradient flow and applying curve evolution techniques. Experimental results on both synthetic and real images confirm the effectiveness of the proposed method. Compared with the previous active contour models formulated to solve the same nonparametric statistical segmentation problem, our method performs well in efficiency and evolution time.

In this paper, we present a novel level set method (LSM) for image segmentation. By utilizing the Bayesian rule, we design a nonlinear adaptive velocity and a probability-weighted stopping force to implement a robust segmentation for objects with weak boundaries. The proposed method is featured by the following three properties: 1) it automatically determines the curve to shrink or expand by utilizing the Bayesian rule to involve the regional features of images; 2) it drives the curve evolve with an appropriate speed to avoid the leakage at weak boundaries; and 3) it reduces the influence of false boundaries, i.e., edges far away from objects of interest. We applied the proposed segmentation method to artificial images, medical images and the BSD-300 image dataset for qualitative and quantitative evaluations. The comparison results show the proposed method performs competitively, compared with the LSM and its representative variants.

Very high resolution (VHR) satellite images provide a mass of detailed information which can be used for urban planning, mapping, security issues, or environmental monitoring. Nevertheless, the processing of this kind of image is timeconsuming, and extracting the needed information from among the huge quantity of data is a real challenge. For some applications such as natural disaster prevention and monitoring (typhoon, flood, bushfire, etc.), the use of fast and effective processing methods is demanded. Furthermore, such methods should be selective in order to extract only the information required to allow an efficient interpretation. For this purpose, we propose a texture region segmentation method using the level set algorithm and the multi-kernel theory. We design a selective and local multi-kernel stop function for which the regularization term depends on the fuzzy membership degree of a given pixel to be on the boundary or not. Favored by its local nature, the method is accelerated by means of an NVIDIA graphics processing unit programming. The new algorithm is selective, effective, and fast. Experimental results on VHR satellite images demonstrate subjectively and objectively the effectiveness of the proposed method.

In the last decades, due to the development of the parallel programming, the lattice Boltzmann method (LBM) has attracted much attention as a fast alternative approach for solving partial differential equations. In this paper, we first designed an energy functional based on the fuzzy c-means objective function which incorporates the bias field that accounts for the intensity inhomogeneity of the real-world image. Using the gradient descent method, we obtained the corresponding level set equation from which we deduce a fuzzy external force for the LBM solver based on the model by Zhao. The method is fast, robust against noise, independent to the position of the initial contour, effective in the presence of intensity inhomogeneity, highly parallelizable and can detect objects with or without edges. Experiments on medical and real-world images demonstrate the performance of the proposed method in terms of speed and efficiency.

Due to its intrinsic nature which allows to easily handle complex shapes and topological changes, the level set method (LSM) has been widely used in image segmentation. Nevertheless, LSM is computationally expensive, which limits its applications in real-time systems. For this purpose, we propose a new level set algorithm, which uses simultaneously edge, region, and 2D histogram information in order to efficiently segment objects of interest in a given scene. The computational complexity of the proposed LSM is greatly reduced by using the highly parallelizable lattice Boltzmann method (LBM) with a body force to solve the level set equation (LSE). The body force is the link with image data and is defined from the proposed LSE. The proposed LSM is then implemented using an NVIDIA graphics processing units to fully take advantage of the LBM local nature. The new algorithm is effective, robust against noise, independent to the initial contour, fast, and highly parallelizable. The edge and region information enable to detect objects with and without edges, and the 2D histogram information enable the effectiveness of the method in a noisy environment. Experimental results on synthetic and real images demonstrate subjectively and objectively the performance of the proposed method.

This paper presents a new image segmentation method that applies an edge-based level set method in a relay fashion. The proposed method segments an image in a series of nested subregions that are automatically created by shrinking the stabilized curves in their previous subregions. The final result is obtained by combining all boundaries detected in these subregions. The proposed method has the following three advantages: 1) It can be automatically executed without human-computer interactions; 2) it applies the edge-based level set method with relay fashion to detect all boundaries; and 3) it automatically obtains a full segmentation without specifying the number of relays in advance. The comparison experiments illustrate that the proposed method performs better than the representative level set methods, and it can obtain similar or better results compared with other popular segmentation algorithms.

This paper presents a new region-based unified tensor level set model for image segmentation. This model introduces a three-order tensor to comprehensively depict features of pixels, e.g., gray value and the local geometrical features, such as orientation and gradient, and then, by defining a weighted distance, we generalized the representative region-based level set method from scalar to tensor. The proposed model has four main advantages compared with the traditional representative method as follows. First, involving the Gaussian filter bank, the model is robust against noise, particularly the salt- and pepper-type noise. Second, considering the local geometrical features, e.g., orientation and gradient, the model pays more attention to boundaries and makes the evolving curve stop more easily at the boundary location. Third, due to the unified tensor pixel representation representing the pixels, the model segments images more accurately and naturally. Fourth, based on a weighted distance definition, the model possesses the capacity to cope with data varying from scalar to vector, then to high-order tensor. We apply the proposed method to synthetic, medical, and natural images, and the result suggests that the proposed method is superior to the available representative region-based level set method.