Boundary-based Methods Research Articles

Interpretable machine learning methods to explain on-farm yield variability of high productivity wheat in Northwest India

The increasing availability of complex, geo-referenced on-farm data demands analytical frameworks that can guide crop management recommendations. Recent developments in interpretable machine learning techniques offer opportunities to use these methods in agronomic studies. Our objectives were two-fold: (1) to assess the performance of different machine learning methods to explain on-farm wheat yield variability in the Northwestern Indo-Gangetic Plains of India, and (2) to identify the most important drivers and interactions explaining wheat yield variability. A suite of fine-tuned machine learning models (ridge and lasso regression, classification and regression trees, k-nearest neighbor, support vector machines, gradient boosting, extreme gradient boosting, and random forest) were statistically compared using the R 2 , root mean square error (RMSE), and mean absolute error (MAE). The best performing model was again fine-tuned using a grid search approach for the bias-variance trade-off. Three post-hoc model agnostic techniques were used to interpret the best performing model: variable importance (a variable was considered “important” if shuffling its values increased or decreased the model error considerably), interaction strength (based on Friedman’s H-statistic), and two-way interaction (i.e., how much of the total variability in wheat yield was explained by a particular two-way interaction). Model outputs were compared against empirical data to contextualize results and provide a blueprint for future analysis in other production systems. Tree-based and decision boundary-based methods outperformed regression-based methods in explaining wheat yield variability. Random forest was the best performing method in terms of goodness-of-fit and model precision and accuracy with RMSE, MAE, and R 2 ranging between 367 and 470 kg ha −1 , 276–345 kg ha −1 , and 0.44–0.63, respectively. Random forest was then used for selection of important variables and interactions. The most important management variables explaining wheat yield variability were nitrogen application rate and crop residue management, whereas the average of monthly cumulative solar radiation during February and March (coinciding with reproductive phase of wheat) was the most important biophysical variable. The effect size of these variables on wheat yield ranged between 227 kg ha −1 for nitrogen application rate to 372 kg ha −1 for cumulative solar radiation during February and March. The effect of important interactions on wheat yield was detected in the data namely the interaction between crop residue management and disease management and, nitrogen application rate and seeding rate. For instance, farmers’ fields with moderate disease incidence yielded 750 kg ha −1 less when crop residues were removed than when crop residues were retained. Similarly, wheat yield response to residue retention was higher under low seed and N application rates. As an inductive research approach, the appropriate application of interpretable machine learning methods can be used to extract agronomically actionable information from large-scale farmer field data. • Data-driven agronomic research requires new analytical and methodological approaches. • Machine learning methods were used to disentangle complex relationships in farmer field data. • Model-agnostic tools were used to derive agronomic interpretations. • Residue management and N application rate were important management variables for wheat yield. • Residue management interacted with other practices to explain wheat yield variability.

A Comprehensive Review on Temporal-Action Proposal Generation.

Temporal-action proposal generation (TAPG) is a well-known pre-processing of temporal-action localization and mainly affects localization performance on untrimmed videos. In recent years, there has been growing interest in proposal generation. Researchers have recently focused on anchor- and boundary-based methods for generating action proposals. The main purpose of this paper is to provide a comprehensive review of temporal-action proposal generation with network architectures and empirical results. The pre-processing step for input data is also discussed for network construction. The content of this paper was obtained from the research literature related to temporal-action proposal generation from 2012 to 2022 for performance evaluation and comparison. From several well-known databases, we used specific keywords to select 71 related studies according to their contributions and evaluation criteria. The contributions and methodologies are summarized and analyzed in a tabular form for each category. The result from state-of-the-art research was further analyzed to show its limitations and challenges for action proposal generation. TAPG performance in average recall ranges from 60% up to 78% in two TAPG benchmarks. In addition, several future potential research directions in this field are suggested based on the current limitations of the related studies.

Elastic Collision Based Dynamic Partitioning Scheme for Hybrid Simulations.

In hybrid simulations, such as the QM/MM approach, the system is partitioned into regions that are treated at different levels of theory. The key question then becomes how to evaluate the interactions between particles on opposite sides of the boundary. One approach is to place the boundary in such a way that particles near the boundary on both sides are of the same type, thus simplifying the evaluation of the interactions. If mobile particles are present, such as solvent molecules, and particles are allowed to cross the boundary, the conservation of energy and atomic forces is problematic unless the computational effort is increased significantly. By preventing particles from crossing the boundary but allowing the boundary to be flexible, an accurate estimate of average thermodynamic properties is obtained in principle as illustrated by the flexible inner region ensemble separator (FIRES) method [C. Rowley and B. Roux, J. Chem. Theory Comput. 2012, 8, 3526]. In FIRES, a harmonic restraint is applied to particles near the boundary. Therefore, it can occur that particle cross the boundary to some extent resulting in anomalies in the particle density. Here, a constraint approach is presented where particles instantaneously scatter from the boundary. This scattering-adapted FIRES (SAFIRES) implementation makes use of a variable-time-step propagation algorithm where the time step is scaled automatically to identify the moment a collision should occur. If the length of the time step is kept constant, this propagator reduces to a regular Langevin dynamics algorithm, and to the velocity Verlet algorithm for conservative dynamics if the friction coefficient is set to zero. Correct average ensemble statistics are obtained as demonstrated in simulations where, for testing purposes, the particles in the two regions are treated at the same level of theory, namely, a homogeneous Lennard-Jones (LJ) liquid and liquid water based on the TIP4P potential function. In order to illustrate this approach in solid-liquid interface simulations, a LJ liquid in contact with the surface of a crystal is also simulated. The simulations using SAFIRES are shown to reproduce the unconstrained reference simulations without significant deviations in the particle density and the dynamics are shown to conserve energy when coupling to the heat bath is turned off.

Mammogram Segmentation Techniques: A Review

There is a significant development in computer-aided detection (CADe) and computer-aided diagnostic (CADx) systems in recent years. This development coincides with the evolution of computing power and the growth of data. The CAD systems support detections and diagnosis of significant diseases, including cancer. Breast cancer is one of the most prevalent cancers influencing women and causing death around the world. Early detection of breast cancer has a significant effect on treatment. The typical CAD system goes through various steps, including image segmentation, feature extraction, and image classification. Image segmentation plays an important role in CAD systems and simplifies further processing. This review explores popular mammogram segmentation techniques. A mammogram is medical imaging which uses a low-dose x-ray system to see inner tissues of the breast. There are many segmentation techniques used to segment medical images. These techniques can be categorized into five main categories: region-based methods, boundary-based methods, atlas-based methods, model-based methods, and deep learning. A ground truth image is needed to measure the performance of the segmentation algorithm. Different performance measurements were used to evaluate the segmentation process, including accuracy, precision, recall, F1 score, Hausdorff Distance, Jaccard, and Dice Index. The research in mammogram segmentation has yielded promising results, but there is room for improvements.

Efficient three-dimensional super-diffusive model for benign brain tumor segmentation

Brain is most complex and central part of the human body. Millions of cells are present in the brain. Brain tumor is extra unwanted cell in the human brain. It is mainly categorized into benign and malignant. Benign tumor cells have very similar characteristics with its surrounding cells. Its accurate detection and segmentation is very challenging task. Image segmentation methods have major contribution in detection and segmentation of these tumor cells. Segmentation methods are either boundary based or region based. These methods use traditional integral-order calculus. It has been observed that these approaches are unable to detect low variational region such as benign tumor. In the present manuscript, fractional diffusion-based benign brain tumor detection and segmentation method is being proposed. It has been observed that the proposed method is able to detect and segment benign brain tumor region more accurately. Higher accuracy has been obtained due to fractional-order derivative. Frequency domain derivative definition has been used in the proposed method due to simplicity and low computational cost. A hardware model of the proposed work has been also presented in the current manuscript. The results obtained have been compared with existing boundary-based and region-based tumor detection and segmentation methods. It has been found that the proposed method is having higher accuracy in benign brain tumor detection and segmentation with low computational cost.

Red blood cell segmentation by thresholding and Canny detector

A computer visualization system that can detect, count and classify cells in bioimages is needed, and the segmentation of cells plays a vital role in this type of system. In the medical field, blood cell testing is among the most important types of clinical examinations. Manual blood cell examination methods using microscopic devices are more time consuming than automatic methods and require radiologists with more technical skills. However, the development of an effective and fully automatic segmentation process for RBCs (red blood cells) remains challenging. This paper proposes an automatic segmentation algorithm for RBCs that automatically computes the threshold image using boundary-based methods after enhancing the local and global details of the output using morphological operations to segment red blood cells in bioimages. The proposed segmentation method exhibited an average accuracy rate of 87.9% in a public human red blood cell dataset. Furthermore, a comparison of this method with gold segmentation and two other methods typically used for this purpose demonstrated that the proposed method was highly robust and outperformed the other methods.

Delineation of Carpal Bones From Hand X-Ray Images Through Prior Model, and Integration of Region-Based and Boundary-Based Segmentations

Image segmentation is critical and challenging in computer vision and medical image analysis. Despite decades of research, existing segmentation algorithms are still subject to typical segmentation problems, such as over-segmentation, under-segmentation, and non-closed and spurious edges. In this paper, taking the carpal bones from hand X-ray images as the foreground regions, we propose a segmentation approach to integrate segmentations from region-based and boundary-based methods to tackle these typical segmentation problems. First, adaptive local thresholding and adaptive Canny edge detection are explored to extract foreground regions and the edge map. Second, the integration of the edge map and foreground regions by XORing is proposed, to tackle the over-segmentation by adding a background boundary from the edge map near the carpal bone boundary so as to break the connection between the foreground and the over-segmented background, to handle under-segmentation by adding a foreground boundary from the edge map near the carpal bone boundary so as to enclose the missing foreground due to under-segmentation, and to complement non-closed edge and spurious edge from the edge map through the carpal bone regions from the local adaptive thresholding. Optionally, marker-controlled watershed segmentation or an active contour-based method is employed to refine the integrated segmentation. The segmented foreground regions are identified through exploring a prior model. The proposed approach has been validated on 30 representative X-ray hand images for non-overlapping carpal bones to yield an average Dice coefficient of 0.976 ± 0.006. The proposed method could provide a theory or tool for accurately segmenting images with discernible boundaries and non-uniform grayscale distribution both within the background and foreground.

Automatic setae segmentation from Chaetoceros microscopic images.

A novel image processing model Grayscale Surface Direction Angle Model (GSDAM) is presented and the algorithm based on GSDAM is developed to segment setae from Chaetoceros microscopic images. The proposed model combines the setae characteristics of the microscopic images with the spatial analysis of image grayscale surface to detect and segment the direction thin and long setae from the low contrast background as well as noise which may make the commonly used segmentation methods invalid. The experimental results show that our algorithm based on GSDAM outperforms the boundary-based and region-based segmentation methods Canny edge detector, iterative threshold selection, Otsu's thresholding, minimum error thresholding, K-means clustering, and marker-controlled watershed on the setae segmentation more accurately and completely.

Covariance-guided One-Class Support Vector Machine

In one-class classification, the low variance directions in the training data carry crucial information to build a good model of the target class. Boundary-based methods like One-Class Support Vector Machine (OSVM) preferentially separates the data from outliers along the large variance directions. On the other hand, retaining only the low variance directions can result in sacrificing some initial properties of the original data and is not desirable, specially in case of limited training samples. This paper introduces a Covariance-guided One-Class Support Vector Machine (COSVM) classification method which emphasizes the low variance projectional directions of the training data without compromising any important characteristics. COSVM improves upon the OSVM method by controlling the direction of the separating hyperplane through incorporation of the estimated covariance matrix from the training data. Our proposed method is a convex optimization problem resulting in one global optimum solution which can be solved efficiently with the help of existing numerical methods. The method also keeps the principal structure of the OSVM method intact, and can be implemented easily with the existing OSVM libraries. Comparative experimental results with contemporary one-class classifiers on numerous artificial and benchmark datasets demonstrate that our method results in significantly better classification performance.

A landscape pattern analysis method based on boundaries and nodes: A case study in upper Minjiang River, China

Traditionally, patch-based analysis at the landscape scale fragmentation has been used in the study of landscape ecology while the study of boundary and node has not been considered as much detail until recently. This study investigated the possibility of applying boundary- and node-based methods in landscape pattern analysis to the upper reaches of the Minjiang River in the southwestern China. Boundary-based and node-based landscape indices were selected to be used in analyzing changes in landscape patterns, and the results were compared with analysis using traditional pattern indices. We compared the responses of patch-area-based, boundary-length-based and node-number-based indices, and concluded that boundary-based and node-based indices are more sensitive to disturbance than patch-based indices with various patterns, and node-based indices are even more sensitive than boundary-based ones. Thus, the results suggest that boundary-based and node-based pattern analysis methods provide helpful supplementary information to traditional patch-based pattern analysis methods. The results about pattern dynamics of landscapes in the upper reaches of the Minjiang River based on boundaries and nodes showed that with human disturbance, the dominance of forest landscape was weakened by other landscape types; thus the landscape pattern of the study area became more homogeneous and the boundary network became more complex. These changes further augmented disturbance interfaces in the landscape and increased the possibilities of further landscape fragmentation.

Oriented Image Foresting Transform Segmentation by Seed Competition

Seed-based methods for region-based image segmentation are known to provide satisfactory results for several applications, being usually easy to extend to multidimensional images. However, while boundary-based methods like live wire can easily incorporate a preferred boundary orientation, regionbased methods are usually conceived for undirected graphs, and do not resolve well between boundaries with opposite orientations. This motivated researchers to investigate extensions for some region-based frameworks, seeking to better solve oriented transitions. In this same spirit, we discuss how to incorporate this orientation information in a region-based approach called “IFT segmentation by seed competition” by exploring digraphs. We give direct proof for the optimality of the proposed extensions in terms of energy functions associated with the cuts. To stress these theoretical results, we also present an experimental evaluation that shows the obtained gains in accuracy for some 2D and 3D data sets of medical images.

Combining Boundary-Based Methods With Tensor-Based Morphometry in the Measurement of Longitudinal Brain Change

Tensor-based morphometry is a powerful tool for automatically computing longitudinal change in brain structure. Because of bias in images and in the algorithm itself, however, a penalty term and inverse consistency are needed to control the over-reporting of nonbiological change. These may force a tradeoff between the intrinsic sensitivity and specificity, potentially leading to an under-reporting of authentic biological change with time. We propose a new method incorporating prior information about tissue boundaries (where biological change is likely to exist) that aims to keep the robustness and specificity contributed by the penalty term and inverse consistency while maintaining localization and sensitivity. Results indicate that this method has improved sensitivity without increased noise. Thus it will have enhanced power to detect differences within normal aging and along the spectrum of cognitive impairment.

A conceptual framework for the spatial analysis of landscape genetic data

Understanding how landscape heterogeneity constrains gene flow and the spread of adaptive genetic variation is important for biological conservation given current global change. However, the integration of population genetics, landscape ecology and spatial statistics remains an interdisciplinary challenge at the levels of concepts and methods. We present a conceptual framework to relate the spatial distribution of genetic variation to the processes of gene flow and adaptation as regulated by spatial heterogeneity of the environment, while explicitly considering the spatial and temporal dynamics of landscapes, organisms and their genes. When selecting the appropriate analytical methods, it is necessary to consider the effects of multiple processes and the nature of population genetic data. Our framework relates key landscape genetics questions to four levels of analysis: (i) node-based methods, which model the spatial distribution of alleles at sampling locations (nodes) from local site characteristics; these methods are suitable for modeling adaptive genetic variation while accounting for the presence of spatial autocorrelation. (ii) Link-based methods, which model the probability of gene flow between two patches (link) and relate neutral molecular marker data to landscape heterogeneity; these methods are suitable for modeling neutral genetic variation but are subject to inferential problems, which may be alleviated by reducing links based on a network model of the population. (iii) Neighborhood-based methods, which model the connectivity of a focal patch with all other patches in its local neighborhood; these methods provide a link to metapopulation theory and landscape connectivity modeling and may allow the integration of node- and link-based information, but applications in landscape genetics are still limited. (iv) Boundary-based methods, which delineate genetically homogeneous populations and infer the location of genetic boundaries; these methods are suitable for testing for barrier effects of landscape features in a hypothesis-testing framework. We conclude that the power to detect the effect of landscape heterogeneity on the spatial distribution of genetic variation can be increased by explicit consideration of underlying assumptions and choice of an appropriate analytical approach depending on the research question.

3D segmentation of SBFSEM images of neuropil by a graphical model over supervoxel boundaries

The segmentation of large volume images of neuropil acquired by serial sectioning electron microscopy is an important step toward the 3D reconstruction of neural circuits. The only cue provided by the data at hand is boundaries between otherwise indistinguishable objects. This indistinguishability, combined with the boundaries becoming very thin or faint in places, makes the large body of work on region-based segmentation methods inapplicable. On the other hand, boundary-based methods that exploit purely local evidence do not reach the extremely high accuracy required by the application domain that cannot tolerate the global topological errors arising from false local decisions. As a consequence, we propose a supervoxel merging method that arrives at its decisions in a non-local fashion, by posing and approximately solving a joint combinatorial optimization problem over all faces between supervoxels. The use of supervoxels allows the extraction of expressive geometric features. These are used by the higher-order potentials in a graphical model that assimilate knowledge about the geometry of neural surfaces by automated training on a gold standard. The scope of this improvement is demonstrated on the benchmark dataset E1088 (Helmstaedter et al., 2011) of 7.5billionvoxels from the inner plexiform layer of rabbit retina. We provide C++ source code for annotation, geometry extraction, training and inference.

A comparative study of boundary-based intelligent sampling approaches for nonlinear optimization

In order to increase the efficiency and accuracy of design optimization, many efforts have been made on studying the metamodel techniques for effectively representing expensive and complex models. The performance of metamodel-based optimization is largely determined by the sampling method. Mainly due to the difficulty of knowing the appropriate sampling size a priori, termed as intelligent or sequential sampling has gained popularity in recent years. In this paper, a new kind of sampling mode, boundary-based strategies are studied. The major characteristic of intelligent methods is that the new sample is generated based on boundary and other specified samples. In order to evaluate the efficiency and accuracy of such kinds of strategies, a comparative study is implemented with mathematical nonlinear test functions. To validate the feasibility of boundary-based sampling methods, the proposed sampling strategy-based metamodeling technique is used to optimize real-world problems. Comparison of model predications and optimization data shows good agreement.

A hybrid framework for 3D medical image segmentation

In this paper we propose a novel hybrid 3D segmentation framework which combines Gibbs models, marching cubes and deformable models. In the framework, first we construct a new Gibbs model whose energy function is defined on a high order clique system. The new model includes both region and boundary information during segmentation. Next we improve the original marching cubes method to construct 3D meshes from Gibbs models’ output. The 3D mesh serves as the initial geometry of the deformable model. Then we deform the deformable model using external image forces so that the model converges to the object surface. We run the Gibbs model and the deformable model recursively by updating the Gibbs model’s parameters using the region and boundary information in the deformable model segmentation result. In our approach, the hybrid combination of region-based methods and boundary-based methods results in improved segmentations of complex structures. The benefit of the methodology is that it produces high quality segmentations of 3D structures using little prior information and minimal user intervention. The modules in this segmentation methodology are developed within the context of the Insight ToolKit (ITK). We present experimental segmentation results of brain tumors and evaluate our method by comparing experimental results with expert manual segmentations. The evaluation results show that the methodology achieves high quality segmentation results with computational efficiency. We also present segmentation results of other clinical objects to illustrate the strength of the methodology as a generic segmentation framework.

Efficient image representation using binary space partitioning trees

Images are traditionally represented by a set of picture elements, or pixels, defined over a discrete two or three-dimensional lattice. Pixel-based representation of images, however, requires a large amount of storage memory and processing power when performing complex image operations. Several image processing, computer vision, and computer graphics applications have resorted to employing more efficient representations of images. Tree-based representations, such as quadtrees and octrees, have received a great deal of attention by researchers in many imaging, vision and graphics areas. Binary Space Partitioning (BSP) trees have been used in computer graphics as an efficient representation of objects with planar faces (i.e., polyhedra). In this dissertation we propose using BSP trees as an efficient representation of images for image compression and geometric transformation applications. The BSP tree approach partitions the image domain, in a recursive manner, using arbitrarily oriented lines. The result of this partitioning is a set of convex (unpartitioned) regions, known as the BSP tree cells. In order to generate an efficient segmentation of the image, the number of cells has to be as small as possible such that the image signal within each cell is smooth. Therefore, the most critical aspect of a BSP tree construction method is the criterion used for selecting the partitioning lines. Here, we introduce a boundary-based approach and an optimization method for selecting the lines of a BSP tree. Under the former approach, the partitionings are based on the boundaries of the objects within the image under consideration. By using an optimization method, the lines are selected based on a Least-Square-Error (LSE) criterion. As demonstrated in this work, both the boundarybased and LSE methods provide efficient segmentation when applied to complex images. In addition, we develop a multiresolution algorithm for constructing a multiresolution BSP tree. In this case, the coarse (high frequency) information of the image is represented by the nodes close to the root node, and the fine (low frequency) details are contained within the nodes close to the leaves of the BSP tree. Based on these BSP tree construction methods, a coding system is proposed for compressing still images. A hierarchical method for coding the partitioning lines of the tree is developed. Meanwhile, the image signal within the unpartitioned regions (cells) is coded using low order polynomials (zero and first order). Moreover, an optimum pruning algorithm is used to reduce the number of bits required for encoding the BSP tree representation while minimizing distortion. By employing this BSP tree-based compression method on typical images (of human faces), high compression ratios in the range of 50-100 were obtained. Unlike other tree-based representations (e.g. quadtrees), and due to the flexible representation of the BSP tree approach (namely, the use of arbitrarily oriented lines), line-preserving transformations can be applied very efficiently to BSP trees without the need for reconstructing the tree under consideration. In here, we show the result of performing affine and perspective projection transformations on the BSP tree representation of images. These geometric manipulations were achieved by performing parametric transformations on the partitioning lines of the tree.