Boundary Map Research Articles

Intelligent X-ray waste detection and classification via X-ray characteristic enhancement and deep learning

Waste management plays an important role in environmental protection and resource recycling. In fact, traditional waste recycling requires a large number of workers to sort manually, which is not only labor-intensive and time-consuming, but also increases the health risk for workers. To further improve the efficiency of waste detection and classification, related scholars have used computer vision technology to achieve automatic waste management. However, existing methods tend to use RGB images as input, requiring workers to open waste bags and disperse waste items in advance. To avoid comprehensive preparation procedures and identify occluded waste items, we use deep learning models to detect and recognize waste items in X-ray images. Compared to RGB images, X-ray images can effectively preserve boundary information and reflect material types. To take advantage of the inherent characteristics of X-ray images, we propose a novel X-ray Characteristic Enhancement Module (XCEM) to improve the performance of X-ray waste detection. XCEM consists of three sub-modules, namely Hybrid Interaction Module (HIM), Multi-scale Material Perception Module (MMPM), and Boundary Guidance Module (BGM). Specifically, HIM not only generates soft and hard boundary maps, but also achieves information interaction and feature enhancement of different boundary maps. Meanwhile, MMPM can effectively capture different multi-scale material features. In addition, BGM uses auxiliary boundary information and attention mechanisms to effectively enhance critical material features. As a plug-and-play module, XCEM can be readily inserted into the most popular detection pipelines. We select several representative one-stage and two-stage methods and perform related experiments on the WIXRay dataset. Experimental results show that our method can achieve consistent performance gains on diverse detection paradigms. This study promotes the development of intelligent waste inspection and recycling.

TSRNet: A Dual-stream Network for Refining 3D Tooth Segmentation.

The field of 3D tooth segmentation has made considerable advances thanks to deep learning, but challenges remain with coarse segmentation boundaries and prediction errors. In this paper, we introduce a novel learnable method to refine coarse results obtained from existing 3D tooth segmentation algorithms. The refinement framework features a dual-stream network called TSRNet (Tooth Segmentation Refinement Network) to rectify defective boundary and distance maps extracted from the coarse segmentation. The boundary map provides explicit boundary information, while the distance map provides gradient information in the form of the shortest geodesic distance between the vertex and the segmentation boundary. Following well-designed rules, the two refined maps are utilized to move the coarse tooth boundaries toward their correct positions through an iterative refinement process. The two-stage refinement method is validated on both 3D tooth and segmentation benchmark datasets. Extensive experiments demonstrate that our method significantly improves upon the coarse results from baseline methods and achieves state-of-the-art performance. (Code will be publicly available at https://github.com/bibi547/TSRNet.git.).

BCUIS‐Net: A breast cancer ultrasound image segmentation network via boundary‐aware and shape feature fusion

AbstractBreast cancer is a highly lethal disease with the highest mortality rate among women worldwide. Breast tumor segmentation from ultrasound images plays a critical role in enabling early detection, leading to a reduction in mortality rates. However, the challenge of ultrasound breast cancer segmentation arises from factors such as indistinct lesion boundaries, noise artifacts, and inhomogeneous intensity distribution within the lesion region. To address the bottlenecks, a novel boundary‐aware shape feature fusion network (BCUIS‐Net) is proposed to segment breast lesion in ultrasound images. Firstly, a boundary‐aware module (BAM) is put forward to accurately localize the ambiguous tumor regions and boundaries by embedding the horizontal and vertical position information into the channel attention. Subsequently, a shape feature fusion (SFF) module is presented to fuse shape features and segmentation features, in order to adaptively extract their complementary features by aggregating contextual information in an attention module. Specifically, the different levels of features from the encoder are up‐sampled to the original image size and fed into the BAM to predict the boundary map. The boundary and decoder‐generated feature maps are thereafter fused by the SFF module to exploit the complementarity between them to correct errors in segmentation and shape features, effectively eliminating false detections and noise in the features to achieve accurate segmentation of pathological regions. Finally, the shape fusion loss is derived from a combination of the binary cross‐entropy loss and the distance map loss to intelligently penalize incorrect predictions and thus improve the attention to boundary locations. The performance of the network is evaluated in two public breast ultrasound datasets. Experimental results verify that the proposed method obtains superior segmentation results and outperforms the most recent state‐of‐the‐art, in which IOU is increased by 2.15% and 2.59% on UDIAT and BUSI, respectively.

On the solvability and the dimension of the solution set of abstract linear boundary value problems and applications to ODE'S

A solvability theory for linear abstract boundary value problems, that is of linear operator equations subject to an additional (“boundary”) condition given by a linear operator, is developed using a new approach based on normal solvability of the induced map or a pseudoinverse of the boundary map when restricted to the null space of the homogeneous problem. Uniqueness and dimension of their solution sets is also established. Applications to BVP's for finite-dimensional as well as infinite-dimensional systems of ordinary differential equations defined on the half line or the whole line are given. Systems having exponential dichotomy, or satisfying Riccati type inequalities are studied, as well as systems with asymptotically hyperbolic linear part and systems between pairs of admissible Banach spaces.

Preparation of Administrative Boundary Map of Arahan Kidul Village, Arahan District , Indramayu Regency, West Java

Preparation of Administrative Boundary Map of Arahan Kidul Village, Arahan District , Indramayu Regency, West Java

RGB depth salient object detection via cross‐modal attention and boundary feature guidance

AbstractRGB depth (RGB‐D) salient object detection (SOD) is a meaningful and challenging task, which has achieved good detection performance in dealing with simple scenes using convolutional neural networks, however, it cannot effectively handle scenes with complex contours of salient objects or similarly coloured salient objects and background. A novel end‐to‐end framework is proposed for RGB‐D SOD, which comprises of four main components: the cross‐modal attention feature enhancement (CMAFE) module, the multi‐level contextual feature interaction (MLCFI) module, the boundary feature extraction (BFE) module, and the multi‐level boundary attention guidance (MLBAG) module. The CMAFE module retains the more effective salient features by employing a dual‐attention mechanism to filter noise from two modalities. In the MLCFI module, a shuffle operation is used for high‐level and low‐level channels to promote cross‐channel information communication, and rich semantic information is extracted. The BFE module converts salient features into boundary features to generate boundary maps. The MLBAG module produces saliency maps by aggregating multi‐level boundary saliency maps to guide cross‐modal features in the decode stage. Extensive experiments are conducted on six public benchmark datasets, with the results demonstrating that the proposed model significantly outperforms 23 state‐of‐the‐art RGB‐D SOD models with regards to multiple evaluation metrics.

Field Patch Extraction Based on High-Resolution Imaging and U2-Net++ Convolutional Neural Networks

Accurate extraction of farmland boundaries is crucial for improving the efficiency of farmland surveys, achieving precise agricultural management, enhancing farmers’ production conditions, protecting the ecological environment, and promoting local economic development. Remote sensing and deep learning are feasible methods for creating large-scale farmland boundary maps. However, existing neural network models have limitations that restrict the accuracy and reliability of agricultural parcel extraction using remote sensing technology. In this study, we used high-resolution satellite images (2 m, 1 m, and 0.8 m) and the U2-Net++ model based on the RSU module, deep separable convolution, and the channel-spatial attention mechanism module to extract different types of fields. Our model exhibited significant improvements in farmland parcel extraction compared with the other models. It achieved an F1-score of 97.13%, which is a 7.36% to 17.63% improvement over older models such as U-Net and FCN and a more than 2% improvement over advanced models such as DeepLabv3+ and U2-Net. These results indicate that U2-Net++ holds the potential for widespread application in the production of large-scale farmland boundary maps.

BGNet: Boundary‐guided network for polyp segmentation

AbstractColonoscopy is one of the most direct and effective methods for detecting colon polyps; they are crucial for early screening and prevention of colorectal cancer (CRC). Accurate segmentation of polyp images is significant for the clinical management and treatment of CRC. However, polyp image segmentation is a challenging task because polyps vary in size, shape, and color, and have low contrast with the surrounding tissue and mucosa. To address these challenges, we propose a novel network called BGNet for polyp segmentation. BGNet consists of three modules: a boundary feature extraction module (BFEM), a mutual optimization module (MOM), and a convolutional attention module (CBAM). The BFEM extracts boundary features and predicts boundary maps, which complement and guide region features to accurately predict polyp regions and help BGNet generate precise prediction masks. The MOM optimizes the extracted features to enhance fine‐grained feature representation. The CBAM performs attention operations on the feature map to focus on informative regions and suppress irrelevant information. We conduct quantitative and qualitative evaluations on five benchmark datasets, and the results show that BGNet outperforms other methods and has strong learning ability and generalization ability.

Entropy current and fluid-gravity duality in Gauss-Bonnet theory

Working within the approximation of small amplitude expansion, recently an entropy current has been constructed on the horizons of dynamical black hole solution in any higher derivative theory of gravity. In this note, we have dualized this horizon entropy current to a boundary entropy current in an asymptotically AdS black hole metric with a dual description in terms of dynamical fluids living on the AdS boundary. This boundary entropy current is constructed using a set of mapping functions relating each point on the horizon to a point on the boundary. We have applied our construction to black holes in Einstein-Gauss-Bonnet theory. We have seen that up to the first order in derivative expansion, Gauss-Bonnet terms do not add any extra corrections to fluid entropy as expected. However, at the second order in derivative expansion, the boundary current will non-trivially depend on how we choose our horizon to boundary map, which need not be expressible entirely in terms of fluid variables. So generically, the boundary entropy current generated by dualizing the horizon current will not admit a fluid dynamical description.

E2EVAP: End-to-end vectorization of smallholder agricultural parcel boundaries from high-resolution remote sensing imagery

Rapid and accurate agricultural parcel mapping from high-resolution remote sensing imagery is fundamental to precision agriculture for smallholder farming systems. However, due to the narrow and small-size parcels, and the significant spatio-spectral variability, the existing two-stage segmentation methods cannot extract individual parcels automatically. In this article, the end-to-end vectorization of smallholder agricultural parcel boundaries (E2EVAP) framework is proposed for extracting the vertices of each parcel boundary individually in smallholder farming regions, where the semantic-contour interaction and topological loss through hierarchical instance representation (called “channel to instance”) are designed for aggregating the foreground features and jointly establishing the topological relationship between instances to alleviate the topological overlap between parcel objects. Vertex shift correction based on the deep attention corner snake module guided by the parcel boundaries is also proposed to adaptively correct the boundary vertex location shift of large-scale irregular parcels. The comprehensive experimental results obtained on the iFLYTEK public agricultural parcel dataset confirm that E2EVAP shows a superior performance (with a mask mAP of 0.335 and a boundary mAP of 0.201), compared with the pixel-based (such as Mask R-CNN, HTC, and SOLOv2) and contour-based (such as DeepSnake and E2EC) benchmark methods. We believe that E2EVAP has the potential to be widely used for accurate vectorization mapping of agricultural parcels in smallholder farming areas such as South Africa and southern China. The code of E2EVAP is at https://github.com/YangPanHZAU/E2EVAP.

Subject-specific whole-brain parcellations of nodes and boundaries are modulated differently under 10 Hz rTMS

Repetitive transcranial magnetic stimulation (rTMS) has gained considerable importance in the treatment of neuropsychiatric disorders, including major depression. However, it is not yet understood how rTMS alters brain’s functional connectivity. Here we report changes in functional connectivity captured by resting state functional magnetic resonance imaging (rsfMRI) within the first hour after 10 Hz rTMS. We apply subject-specific parcellation schemes to detect changes (1) in network nodes, where the strongest functional connectivity of regions is observed, and (2) in network boundaries, where functional transitions between regions occur. We use support vector machine (SVM), a widely used machine learning algorithm that is robust and effective, for the classification and characterization of time intervals of changes in node and boundary maps. Our results reveal that changes in connectivity at the boundaries are slower and more complex than in those observed in the nodes, but of similar magnitude according to accuracy confidence intervals. These results were strongest in the posterior cingulate cortex and precuneus. As network boundaries are indeed under-investigated in comparison to nodes in connectomics research, our results highlight their contribution to functional adjustments to rTMS.

Auto-parametric resonance of flexible viscoelastic beams under interaction between longitudinal and transverse modes

For a system encompassing subsystems, dynamic modal interaction may experience unanticipated vibrations among subsystems that are excited indirectly. This paper focuses attention to a nonlinear auto-parametric resonance phenomenon of a flexible beam introduced via modal coupling of longitudinal forced resonance and transverse parametric resonance under longitudinal pulsatile excitation. The coupled resonance equations are first modeled based on the Euler-Bernoulli beam theory by incorporating nonlinear curvature, von Kármán's nonlinear strain-displacement relationship and the flexible beam model characterizing nonlinear inertia and displacement interaction. The Galerkin truncation-incremental harmonic balance method (IHBM) is then introduced to obtain both the stability boundary and amplitude-frequency bifurcation maps. Through the corresponding relationship between stability boundary and bifurcation diagram, the energy exchange between two modes is identified as well as the instability region and bifurcation topology can be tuned. It is found that the modal interaction greatly widens the instability region with at most two resonance bands of lateral resonance. By tuning structural and foundation damping, resonance can be controlled to occur only in one frequency band of the pair as desired. This research provides guidance to design broadband energy harvesting or vibration suppression equipment with tunable working frequency band for using the strategy of auto-parametric resonance.

DDL-MVS: Depth Discontinuity Learning for Multi-View Stereo Networks

We propose an enhancement module called depth discontinuity learning (DDL) for learning-based multi-view stereo (MVS) methods. Traditional methods are known for their accuracy but struggle with completeness. While recent learning-based methods have improved completeness at the cost of accuracy, our DDL approach aims to improve accuracy while retaining completeness in the reconstruction process. To achieve this, we introduce the joint estimation of depth and boundary maps, where the boundary maps are explicitly utilized for further refinement of the depth maps. We validate our idea by integrating it into an existing learning-based MVS pipeline where the reconstruction depends on high-quality depth map estimation. Extensive experiments on various datasets, namely DTU, ETH3D, “Tanks and Temples”, and BlendedMVS, show that our method improves reconstruction quality compared to our baseline, Patchmatchnet. Our ablation study demonstrates that incorporating the proposed DDL significantly reduces the depth map error, for instance, by more than 30% on the DTU dataset, and leads to improved depth map quality in both smooth and boundary regions. Additionally, our qualitative analysis has shown that the reconstructed point cloud exhibits enhanced quality without any significant compromise on completeness. Finally, the experiments reveal that our proposed model and strategies exhibit strong generalization capabilities across the various datasets.

Accurate polyp segmentation through enhancing feature fusion and boosting boundary performance

Colorectal polyps are known to be potential precursors to colorectal cancer. Effective polyp segmentation during colonoscopy examinations can help the clinicians accurately locate potential polyp areas, and reduce misdiagnosis and missed diagnosis. Although existing approaches have achieved significant breakthroughs in medical image segmentation, polyp segmentation is still far from being solved. This is mainly due to the following reasons: (1) most of them tend to ignore the feature misalignment issues during the feature aggregation process; and (2) few algorithms explicitly consider the impact of boundary information on the performance of polyp segmentation. To solve the above issues, we formulate a novel neural network for polyp segmentation in endoscopy images. Different from existing approaches, we explore a new paradigm to enhance multi-level feature fusion by introducing the Feature Fusion Module, which leverages the semantic offset field learned to align the multi-level feature maps for resolving the feature misalignment issue. In addition, we design an auxiliary boundary branch to focus on boundary-aware information, and thus boost the performance of boundary prediction. Specifically, the reference boundary map learned through end-to-end optimization can be considered as a complementary feature of the high-level semantics representation, and then integrated into the main branch via the Boundary Embedding Module, which in turn promotes further refinement of the prediction, especially the boundary. Extensive experiments demonstrate that our approach achieves state-of-the-art performance in polyp segmentation.

Flow instability of supercritical carbon dioxide in Converging–Diverging tube

In promising energy conversion system of supercritical carbon dioxide (SCO2), flow instabilities easily occur in this system when it operates near the critical pressure and pseudo-critical temperature due to drastic change of the thermal properties. The converging–diverging tube (CDT) is widely used in enhance heat transfer because of its unique advantage. However, there is no research on instability, and the mechanism of density wave oscillations (DWOs) in CDT remains unclear. Accordingly, in this paper, we studied the instability mechanism and occurrence conditions of DWOs for CDT. We found that the critical trans-pseudocritical number Ntpc of the CDT was reduced by 20.6% compared to that of the straight tube due to the feedback delay effect. This was because the frictional resistance increases per unit mass of flow led to a decreasing flow velocity. The effects of the flow direction on the DWOs in the CDT and straight tubes were different due to the buoyancy changing. The critical Ntpc changed little at different pressures due to the small differences of the average value of βPCp. Finally, we obtained the dimensionless instability boundary map of the CDT to distinguish and identify the regions where instability occurs. This work helps clarify the mechanism of DWOs, identifies the differences between the DWOs in CDTs and straight pipes, and guides the safe operation of CDTs.

Visual Interpretation of Machine Learning: Genetical Classification of Apatite from Various Ore Sources

Machine learning provides solutions to a diverse range of problems in high-dimensional datasets in geosciences. However, machine learning is generally criticized for being an enigmatic black box as it focusses on results but ignores the processes. To address this issue, we used supervised decision boundary maps (SDBM) to visually illustrate and interpret the machine learning process. We constructed a SDBM to classify the ore genetics from 1551 trace element data of apatite in various types of deposits. Attribute-based visual explanation of multidimensional projections (A-MPs) was introduced to SDBM to further demonstrate the correlation between features and machine learning process. Our results show that SDBM explores the interpretability of machine learning process and the A-MPs approach reveals the role of trace elements in machine learning classification. Combining SDBM and A-MPs methods, we propose intuitive and accurate discrimination diagrams and the most indicative elements for ore genetic types. Our work provides novel insights for the visualization application of geo-machine learning, which is expected to be a powerful tool for high-dimensional geochemical data analysis and mineral deposit exploration.

Boundary-oriented Network for Automatic Breast Tumor Segmentation in Ultrasound Images.

Breast cancer is considered as the most prevalent cancer. Using ultrasound images is a momentous clinical diagnosis method to locate breast tumors. However, accurate segmentation of breast tumors remains an open problem due to ultrasound artifacts, low contrast, and complicated tumor shapes in ultrasound images. To address this issue, we proposed a boundary-oriented network (BO-Net) for boosting breast tumor segmentation in ultrasound images. The BO-Net boosts tumor segmentation performance from two perspectives. Firstly, a boundary-oriented module (BOM) was designed to capture the weak boundaries of breast tumors by learning additional breast tumor boundary maps. Second, we focus on enhanced feature extraction, which takes advantage of the Atrous Spatial Pyramid Pooling (ASPP) module and Squeeze-and-Excitation (SE) block to obtain multi-scale and efficient feature information. We evaluate our network on two public datasets: Dataset B and BUSI. For the Dataset B, our network achieves 0.8685 in Dice, 0.7846 in Jaccard, 0.8604 in Precision, 0.9078 in Recall, and 0.9928 in Specificity. For the BUSI dataset, our network achieves 0.7954 in Dice, 0.7033 in Jaccard, 0.8275 in Precision, 0.8251 in Recall, and 0.9814 in Specificity. Experimental results show that BO-Net outperforms the state-of-the-art segmentation methods for breast tumor segmentation in ultrasound images. It demonstrates that focusing on boundary and feature enhancement creates more efficient and robust breast tumor segmentation.

The microstructure and mechanical properties of 316L austenitic stainless steel prepared by forge and laser melting deposition

In this study, the 316L austenitic stainless steel blocks were prepared by forge and laser melting deposition (LMD). To clarify the effect of deposition strategy on microstructure and mechanical properties of 316L samples, three different deposition strategies (30°, 67° and 90°) are selected. The results showed that the molten pool of LMDed sample consists of a large number of coarse columnar grains and a few irregular equiaxed grains, which contains plentiful cellular and fiber sub-structures with an average size of ∼4 μm. The LMD process can significantly improve the comprehensive mechanical properties perpendicular to the build direction of 316L austenitic stainless steel, which was mainly attributed to sub-structural strengthening, but also slightly due to the local misorientation and texture distribution. The yield strength and ultimate tensile strength of R90 sample increased to 484.0 MPa and 627.6 MPa from 305.3 MPa to 601.4 MPa of as-forged 316L, and its elongation has a negligible reduction. The sub-structures produced during solidification have the same fine grain strengthening effect as the high-angle grain boundary, and the yield strength and average size of sub-structures conform to the Hall-Petch relationship. The deterioration of the plastic deformation capacity of the R67 sample is attributed to the larger thermal stress and texture distribution with a certain hard orientation. The plastic deformation of as-forged sample is mainly coordinated by multi-system slip, while twinning and slip are activated together during deformation of LMDed 316L austenitic stainless steel. The grain boundary map indicated that the strain location preferentially appears along cell wall of sub-structures, which promotes the formation of sub-grain boundaries during deformation of LMDed samples.

Stability Analysis of Supervised Decision Boundary Maps

Stability Analysis of Supervised Decision Boundary Maps

A Machine Learning Framework for Assessing Urban Growth of Cities and Suitability Analysis

Rural–urban immigration, regional wars, refugees, and natural disasters all bring to prominence the importance of studying urban growth. Increased urban growth rates are becoming a global phenomenon creating stress on agricultural land, spreading pollution, accelerating global warming, and increasing water run-off, which adds exponentially to pressure on natural resources and impacts climate change. Based on the integration of machine learning (ML) and geographic information system (GIS), we employed a framework to delineate future urban boundaries for future expansion and urban agglomerations. We developed it based on a Time Delay Neural Network (TDNN) that depends on equal time intervals of urban growth. Such an approach is used for the first time in urban growth as a predictive tool and is coupled with Land Suitability Analysis, which incorporates both qualitative and quantitative data to propose evaluated urban growth in the Greater Irbid Municipality, Jordan. The results show the recommended future spatial expansion and proposed results for the year 2025. The results show that urban growth is more prevalent in the eastern, northern, and southern areas and less in the west. The urban growth boundary map illustrates that the continuation of urban growth in these areas will slowly further encroach upon and diminish agricultural land. By means of suitability analysis, the results showed that 51% of the region is unsuitable for growth, 43% is moderately suitable and only 6% is suitable for growth. Based on TDNN methodology, which is an ML framework that is dependent on the growth of urban boundaries, we can track and predict the trend of urban spatial expansion and thus develop policies for protecting ecological and agricultural lands and optimizing and directing urban growth.