Automatic Decomposition Research Articles

Automatic decomposition of 3D solids into contractible pieces using Reeb graphs

This work is motivated by the need to generate volumetric spline models for isogeometric analysis. There exist numerous constructions of volumetric spline models that represent contractible solids. We present a novel decomposition algorithm that splits general solids into pieces that can be dealt with by these existing methods. More precisely, we present a method to automatically decompose solid objects in boundary representation into pieces with fewer or no tunnels by cutting them with auxiliary surfaces. The segmentation is guided by a reduced form of the object’s boundary and volume Reeb graphs with respect to several Morse functions, the level sets of which define the cutting surfaces. Special attention is paid to the selection of suitable cutting surfaces, where we employ a quality criterion to avoid the creation of badly shaped pieces.

Evaluation of user-guided semi-automatic decomposition tool for hexahedral mesh generation

Abstract Volumetric decomposition is essential for all-hexahedral mesh generation. Because fully automatic decomposition methods that can generate high-quality hexahedral meshes for arbitrary volumes have yet to be realized, manual decomposition is still required frequently. Manual decomposition is a laborious process and requires a high level of user expertise. Therefore, a user-guided semi-automatic tool to reduce the human effort and lower the requirement of expertise is necessary. To date, only a few of these approaches have been proposed, and a lack of user evaluation makes it difficult to improve upon this approach. Based on our previous work, we present a user evaluation of a user-guided semi-automatic tool that provides visual guidance to assist users in determining decomposition solutions, accepts sketch-based inputs to create decomposition surfaces, and simplifies the decomposition commands. This user evaluation investigated (1) the usability of the visual guidance, (2) the types of visual guidance essential for decomposition, (3) the effectiveness of the sketch-based decomposition, and (4) the performance differences between beginner and experienced users using the sketch-based decomposition. The result and user feedback indicate that the tool enables users who have limited prior experience or familiarity with the computer-aided engineering software to perform volumetric decomposition more efficiently. The visual guidance increases the success rate of the user's decomposition solution by 28%. The sketch-based decomposition significantly reduces 46% of the user's time on creating decomposition surfaces and setting up decomposition commands. Highlights Evaluations of the user-guided semi-automatic decomposition tool were performed. The visual guidance lowers the required user's expertise level. Sweeping schemes and sweepable regions are essential visual guidance to guide users. The sketch-based decomposition significantly reduces time cost and human efforts.

An algorithm for decomposition of sub-domains and quadrilateral mesh generation with line constraints

The technique for quadrilateral mesh generation on arbitrary 2-D domain has been quite mature; however, the existing methods and software can not directly deal with mesh generation with internal line constraints. In flood analysis and other analysis, the analysis models often contain a large number of constraint lines. In mesh generation, the sides of elements besides constraint lines must be attached to constraint lines, so that boundary conditions can be defined on these constraint lines. In this paper, a general method for automatic quadrilateral mesh generation with internal line constraints is presented. The mesh generation process is mainly divided into two stages, i.e. the automatic decomposition of sub-domains and quadrilateral mesh generation in each sub-domain. The sub-domains enclosed by constraint lines and boundaries are determined automatically with the proposal algorithm in this paper. Then nodes are generated on each edge of sub-domains based on mesh size requirements. By solving an integer linear programming problem, the number of nodes on each edge is adjusted automatically, so that the total number of boundary nodes in each sub-domain is even, satisfying the necessary condition for generating all-quadrilateral mesh. For free constraint lines in each sub-domain, the method of regarding constraint lines as holes with zero area is proposed, so that constraint lines can be treated as internal boundaries. The merging method of constraint lines and outer boundary is given, which can deal with all possible distribution of constraint lines and has generality. The examples of mesh generation with a large number of constraint lines are given to demonstrate the reliability of proposed method.

Motor Units Discriminated From High-density, Surface Electromyography Require Careful Scrutiny.

High-density, surface electromyography (EMG) allows concurrent recording of motor unit (MU) activity by increasing the recording area and the number of recording sites. The recorded signal comprises overlapping MU action potentials, some of which can be identified with a decomposition algorithm. However, such analyses often produce results that contradict the classical findings on motor unit discharge characteristics. PURPOSE: To evaluate the quality of the interspike intervals (ISIs) extracted from a high-density surface EMG recording by a decomposition algorithm. METHODS: Muscle activity was recorded using high-density, surface EMG (4x8 grid with a 10 mm interelectrode distance) in 10 persons with multiple sclerosis during steady, isometric dorsiflexor contractions on their less-affected limb. The target force was 10% of maximum. Custom Matlab software was used to perform a convolution kernel compensation model and a blind-source separation algorithm to discriminate individual MUs from EMG recordings. The discharge times (DT) from the discriminated MUs were used to calculate the ISIs (ISI = DTn-DTn-1). For each MU, the average ISI, average coefficient of variation for ISI (ISICV), kurtosis for the distribution of ISIs, and skewness for the distribution of ISIs was calculated. ISIs outside of the range of 25 ms - 400 ms were deemed inaccurate and removed from the data and the same analyses were repeated. Data are reported as mean [95% CI]. RESULTS: The ISIs (n = 33,980; 303 [285-321] per MU) of 113 MUs were examined. Before removing ISIs that did not meet the inclusion criteria (25 - 400 ms), the average ISI was 136 ms [123 - 151 ms], the average ISICV was 72% [50 - 95%], the average kurtosis was 53.5 [39.1 - 67.8], and the average skewness was 4.4 [3.5 - 4.3]. On average, 3.34% (range 0 - 31%) of the ISIs per MU were outside the acceptable range. After ISI removal, the average ISI was 120 ms [114 - 125 ms], the average ISICV was 28% [26 - 31%], the average kurtosis was 11.0 [8.9 - 13.2], and the average skewness was 1.95 [1.70 - 2.20]. Most removed ISIs were <25 ms. CONCLUSION: The findings quantify the extent to which MU discharge times extracted from high-density, surface EMG recordings with an automatic decomposition algorithm can be confounded by discrimination errors.

A Real-Time Method for Decoding the Neural Drive to Muscles Using Single-Channel Intra-Muscular EMG Recordings.

The neural command from motor neurons to muscles - sometimes referred to as the neural drive to muscle - can be identified by decomposition of electromyographic (EMG) signals. This approach can be used for inferring the voluntary commands in neural interfaces in patients with limb amputations. This paper proposes for the first time an innovative method for fully automatic and real-time intramuscular EMG (iEMG) decomposition. The method is based on online single-pass density-based clustering and adaptive classification of bivariate features, using the concept of potential measure. No attempt was made to resolve superimposed motor unit action potentials. The proposed algorithm was validated on sets of simulated and experimental iEMG signals. Signals were recorded from the biceps femoris long-head, vastus medialis and lateralis and tibialis anterior muscles during low-to-moderate isometric constant-force and linearly-varying force contractions. The average number of missed, duplicated and erroneous clusters for the examined signals was [Formula: see text], [Formula: see text], and [Formula: see text], respectively. The average decomposition accuracy (defined similar to signal detection theory but without using True Negatives in the denominator) and coefficient of determination (variance accounted for) for the cumulative discharge rate estimation were [Formula: see text], and [Formula: see text], respectively. The time cost for processing each 200[Formula: see text]ms iEMG interval was [Formula: see text] (21-97)[Formula: see text]ms. However, computational time generally increases over time as a function of frames/signal epochs. Meanwhile, the incremental accuracy defined as the accuracy of real-time analysis of each signal epoch, was [Formula: see text]% for epochs recorded after initial one second. The proposed algorithm is thus a promising new tool for neural decoding in the next-generation of prosthetic control.

Automatic Logic-Based Benders Decomposition with MiniZinc

Logic-based Benders decomposition (LBBD) is a powerful hybrid optimisation technique that can combine the strong dual bounds of mixed integer programming (MIP) with the combinatorial search strengths of constraint programming (CP). A major drawback of LBBD is that it is a far more involved process to implement an LBBD solution to a problem than the "model-and-run" approach provided by both CP and MIP. We propose an automated approach that accepts an arbitrary MiniZinc model and solves it using LBBD with no additional intervention on the part of the modeller. The design of this approach also reveals an interesting duality between LBBD and large neighborhood search (LNS). We compare our implementation of this approach to CP and MIP solvers on 4 different problem classes where LBBD has been applied before.

Parallel 2D Ray Casting Algorithm for Brain Cell Registration with Brodmann’s Layer Decomposition from Nissl-stained Mouse Cortex Images

In this article we describe some methods for automatic cortex layers decomposition and cortex cell registration. For Brodmann’s layers decomposition, we use special density function calculation. Cell registration process consists of two steps: cell center detection and cell segmentation. To detect cell centers, we use an algorithm based on morphological operations. For cell segmentation, we use the inner ray casting method involving the calculation of intersections with Canny edges. We also propose a parallel version of our methods. Parallel implementation was developed using OpenMP. Performance characteristics of the implementation are also presented

Decomposition of business process models into reusable sub-diagrams

In this paper, an approach to automatic decomposition of business process models is proposed. According to our method, an existing BPMN diagram is disassembled into reusable parts containing the desired number of elements. Such elements and structure can work as design patterns and be validated by a user in terms of correctness. In the next step, these component models are categorised considering their parameters such as resources used, as well as input and output data. The classified components may be considered a repository of reusable parts, that can be further applied in the design of new models. The proposed technique may play a significant role in facilitating the business process redesign procedure, which is of a great importance regarding engineering and industrial applications.

Dominant modal decomposition method

The paper deals with the automatic decomposition of experimental frequency response functions (FRF׳s) of mechanical structures. The decomposition of FRF׳s is based on the Green function representation of free vibratory systems. After the determination of the impulse dynamic subspace, the system matrix is formulated and the poles are calculated directly. By means of the corresponding eigenvectors, the contribution of each element of the impulse dynamic subspace is determined and the sufficient decomposition of the corresponding FRF is carried out. With the presented dominant modal decomposition (DMD) method, the mode shapes, the modal participation vectors and the modal scaling factors are identified using the decomposed FRF׳s. Analytical example is presented along with experimental case studies taken from machine tool industry.

Higher‐order conformal decomposition finite element method

AbstractNumerical simulations of structures using higher‐order finite elements is still a challenging task, in particular for domains with curved boundaries. A new higher‐order accurate approach is proposed, combining the advantages of the classical p‐FEM with embedded domain methods. Boundaries and/or interfaces are described implicitly using the level set method. In the elements cut by the zero level set, an automatic decomposition into interface aligned, i. e. conforming sub‐elements is realized. Transfinite mappings are utilized to construct higher‐order sub‐elements by mappings of reference elements to the two sides of the boundary or interface. It is shown that although the resulting sub‐elements are not always well‐shaped, optimal convergence rates are possible. (© 2016 Wiley‐VCH Verlag GmbH &amp; Co. KGaA, Weinheim)

Preparation of Poly Acrylic Acid Acrylamide Nanometer Material and its Drug Delivery Control Potential in Alkaline Condition

In order to explore a smart carrier which can control drug delivery in special pH environment, a type of nanometer material, namely, poly acrylic acid acrylamide (P(AA-AM)) was synthesized with acrylic acid and acrylamide as monomers. Its structure was determined using Fourier transform infrared spectroscopy (FTIR) and its swelling properties were tested by measuring its weight evolution. The drug delivery control potential was estimated by monitoring its swelling performances in the solutions of different pH values. The cycle characteristics were also investigated to determine its degradation properties in the environment. The results indicate that the synthesized material is sensitive to pH and is suitable for controlling drug delivery at the pH range from 10 to11 and below 7. Furthermore, this material can be automatic decomposition in the environment so that it is a degradable green material without causing burden to the environment.

Conformal higher‐order remeshing schemes for implicitly defined interface problems

SummaryA new higher‐order accurate method is proposed that combines the advantages of the classical p‐version of the FEM on body‐fitted meshes with embedded domain methods. A background mesh composed by higher‐order Lagrange elements is used. Boundaries and interfaces are described implicitly by the level set method and are within elements. In the elements cut by the boundaries or interfaces, an automatic decomposition into higher‐order accurate sub‐elements is realized. Therefore, the zero level sets are detected and meshed in a first step, which is called reconstruction. Then, based on the topological situation in the cut element, higher‐order sub‐elements are mapped to the two sides of the boundary or interface. The quality of the reconstruction and the mapping largely determines the properties of the resulting, automatically generated conforming mesh. It is found that optimal convergence rates are possible although the resulting sub‐elements are not always well‐shaped. Copyright © 2016 John Wiley &amp; Sons, Ltd.

Computation of Galerkin Double Surface Integrals in the 3-D Boundary Element Method

Many boundary element integral equation kernels are based on the Green's functions of the Laplace and Helmholtz equations in three dimensions. These include, for example, the Laplace, Helmholtz, elasticity, Stokes, and Maxwell's equations. Integral equation formulations lead to more compact, but dense linear systems. These dense systems are often solved iteratively via Krylov subspace methods, which may be accelerated via the fast multipole method. There are advantages to Galerkin formulations for such integral equations, as they treat problems associated with kernel singularity, and lead to symmetric and better conditioned matrices. However, the Galerkin method requires each entry in the system matrix to be created via the computation of a double surface integral over one or more pairs of triangles. There are a number of semi-analytical methods to treat these integrals, which all have some issues, and are discussed in this paper. We present novel methods to compute all the integrals that arise in Galerkin formulations involving kernels based on the Laplace and Helmholtz Green's functions to any specified accuracy. Integrals involving completely geometrically separated triangles are non-singular and are computed using a technique based on spherical harmonics and multipole expansions and translations, which results in the integration of polynomial functions over the triangles. Integrals involving cases where the triangles have common vertices, edges, or are coincident are treated via scaling and symmetry arguments, combined with automatic recursive geometric decomposition of the integrals. Example results are presented, and the developed software is available as open source.

An Empirical Study on Improving the Speed and Generalization of Neural Networks Using a Parallel Circuit Approach

One of the common problems of neural networks, especially those with many layers, consists of their lengthy training time. We attempt to solve this problem at the algorithmic level, proposing a simple parallel design which is inspired by the parallel circuits found in the human retina. To avoid large matrix calculations, we split the original network vertically into parallel circuits and let the backpropagation algorithm flow in each subnetwork independently. Experimental results have shown the speed advantage of the proposed approach but also point out that this advantage is affected by multiple dependencies. The results also suggest that parallel circuits improve the generalization ability of neural networks presumably due to automatic problem decomposition. By studying network sparsity, we partly justified this theory and proposed a potential method for improving the design.

Intrinsic image estimation using near- $$L_0$$ L 0 sparse optimization

The objective of intrinsic images estimation is to decompose an input image into its intrinsic shading and reflectance components. This is a well-known under-constrained problem that has long been an open challenge. This paper proposes a novel approach for automatic intrinsic images decomposition that uses a new reflectance sparsity prior. On the basis of the observation that the reflectance of natural objects is commonly piecewise constant, we formalize this constraint on the entire reflectance image using the $$L_0$$L0 sparse loss function that enforces the variation in reflectance images to be of high-frequency and sparse. This new sparsity constraint significantly improves the quality of Retinex intrinsic images estimation. It also functions effectively by combining a class of global sparsity priors on reflectance. Experimental results on MIT benchmark dataset as well as various real-world images and synthetic images demonstrate the effectiveness and versatility of our approach.

Automatic Decomposition of Complex thin Walled CAD Models for Hexahedral Dominant Meshing

This paper describes an automatic method for identifying thin-sheet regions (regions with large lateral dimensions relative to the thickness) for complex thin walled components, with a view to using this information to guide the hex meshing process. This fully automated method has been implemented in a commercial CAD system (Siemens NX) and is based on the interrogation and manipulation of face pairs, which are sets of opposing faces bounding thin-sheet regions. Careful consideration is given to the mapping, merging and intersection of face pairs to generate topologies suitable for sweep meshing thin-sheet regions, and for treating the junctions between adjacent thin-sheet regions. The quality of the resulting hexahedral mesh is considered when making decisions on the generation and positioning of the cutting surfaces required to isolate thin-sheet regions. The resulting decomposition delivers a substantial step towards automatic hexahedral meshing for complex thin-walled geometries. It is proposed that hexahedral meshes be applied to the identified thin-sheet regions by quad meshing one of the faces bounding the thin-sheet region and sweeping it through the thickness to create hexahedral elements.

New advances in visual computing for intelligent processing of visual media and augmented reality

Visual imagery constitutes the most important sensory information for humans. The entire field of acquiring, analyzing and synthesizing visual data by means of computers is called visual computing. It has an extraordinarily wide range of applications, including for example: industrial quality control, street view and driver assistance systems, robot navigation, multimedia systems, and computer games. Visual computing comprises four key areas: computer vision and image processing, computer graphics, virtual and augmented reality, and visualization. It requires deep, interdisciplinary scientific knowledge, in particular in computer science, mathematics, physics, engineering, and cognitive sciences. It tackles high level tasks such as editing and composition of visual content [1] and recognition of sematic content [2–5], as well as basic low level problems such as denoising [6,7] and decomposing [8,9] images, video and 3D shapes. Denosing is often a first step to provide high quality inputs to more complicated tasks such as panoramic image stitching to construct a street view database, and image understanding in computer vision. The object of image denoising is to reduce the noise level, while preserving edges and textures as much as possible. The necessary processing may be done in the image domain, or frequency domain using FFT or wavelets. Recent studies have considered how to represent contour information in images using ridgelets, curvelets and contourlets, as well as finding suitable threshold schemes to remove noise [6]. There is also corresponding work on mesh denosing, which finds structures in terms of positional and normal features [7]. Image decomposition concerns the splitting of an image into two or more components. A fundamental goal in image analysis and computer vision is to extract meaningful components from an image, for tasks such as scene understanding, generation of visual media and many intelligent applications of visual content. Various image decomposition strategies exist. One approach is cartoon and texture decomposition: ref. [9] gave a good survey on the existing decomposition models and extended the nonlinear filter method to decompose an image into three components: the cartoon component, i.e. the main geometric structures, the oscillatory component, or texture, and noise. An image can also be decomposed into a lighting image and a reflectance image known as the intrinsic image. Automatic intrinsic image decomposition remains a significant challenge, particularly for real-world scenes. Recent advances to this longstanding problem are data-driven methods based on large scale datasets of ground truth data. For example, ref. [8] built a dataset of more than 5000 labeled intrinsic images as a basis for a decomposition algorithm, as demonstrated in Figure 1. Computer understanding of scenes or objects within them is a fundamental problem in computer vision, and is crucial to intelligent applications based on visual content. For example, efficient recognition of objects and scene understanding of a live video captured by cameras on a car allows a driver assistance system to instantly feed appropriate information to the driver, or to the cars controls directly.

WrapIt

Wire wrapping is a traditional form of handmade jewelry that involves bending metal wire to create intricate shapes. The technique appeals to novices and casual crafters because of its low cost, accessibility and unique aesthetic. We present a computational design tool that addresses the two main challenges of creating 2D wire-wrapped jewelry: decomposing an input drawing into a set of wires, and bending the wires to give them shape. Our main contribution is an automatic wire decomposition algorithm that segments a drawing into a small number of wires based on aesthetic and fabrication principles. We formulate the task as a constrained graph labeling problem and present a stochastic optimization approach that produces good results for a variety of inputs. Given a decomposition, our system generates a 3D-printed custom support structure, or jig , that helps users bend the wire into the appropriate shape. We validated our wire decomposition algorithm against existing wire-wrapped designs, and used our end-to-end system to create new jewelry from clipart drawings. We also evaluated our approach with novice users, who were able to create various pieces of jewelry in less than half an hour.

Estimating reflex responses in large populations of motor units by decomposition of the high-density surface electromyogram.

Reflex responses of single motor units have been used for the study of spinal circuitries but the methods employed are invasive and limited to the assessment of a relatively small number of motor units. We propose a new approach to investigate reflexes on individual motor units based on high-density surface electromyography (HDsEMG) decomposition. The decomposition of HDsEMG has been previously validated in voluntary isometric contractions but never during reflex activities. The use of HDsEMG decomposition for reflex studies at the individual motor unit level, during constant force contractions, with excitatory and inhibitory stimuli, was validated here by the comparison of results with concurrently recorded intramuscular EMG signals. The validation results showed that HDsEMG decomposition allows an accurate quantification of reflex responses for a large number of individual motor units non-invasively, for both excitatory and inhibitory stimuli. We propose and validate a non-invasive method that enables accurate detection of the discharge times of a relatively large number of motor units during excitatory and inhibitory reflex stimulations. High-density surface electromyography (HDsEMG) and intramuscular EMG (iEMG) were recorded from the tibialis anterior muscle during ankle dorsiflexions performed at 5%, 10% and 20% of the maximum voluntary contraction (MVC) force, in nine healthy subjects. The tibial nerve (inhibitory reflex) and the peroneal nerve (excitatory reflex) were stimulated with constant current stimuli. In total, 416 motor units were identified from the automatic decomposition of the HDsEMG. The iEMG was decomposed using a state-of-the-art decomposition tool and provided 84 motor units (average of two recording sites). The reflex responses of the detected motor units were analysed using the peri-stimulus time histogram (PSTH) and the peri-stimulus frequencygram (PSF). The reflex responses of the common motor units identified concurrently from the HDsEMG and the iEMG signals showed an average disagreement (the difference between number of observed spikes in each bin relative to the mean) of 8.2±2.2% (5% MVC), 6.8±1.0% (10% MVC) and 7.5±2.2% (20% MVC), for reflex inhibition, and 6.5±4.1%, 12.0±1.8% and 13.9±2.4%, for reflex excitation. There was no significant difference between the characteristics of the reflex responses, such as latency, amplitude and duration, for the motor units identified by both techniques. Finally, reflex responses could be identified at higher force (4 of the 9 subjects performed contraction up to 50% MVC) using HDsEMG but not iEMG, because of the difficulty in decomposing the iEMG at high forces. In conclusion, single motor unit reflex responses can be estimated accurately and non-invasively in relatively large populations of motor units using HDsEMG. This non-invasive approach may enable a more thorough investigation of the synaptic input distribution on active motor units at various force levels.

The relationship between morphological awareness and morphological decomposition among English language learners

Morphological awareness facilitates many reading processes. For this reason, L1 and L2 learners of English are often directly taught to use their knowledge of English morphology as a useful reading strategy for determining parts of speech and meaning of novel words. Over time, use of morphological awareness skills while reading develops into an automatic process for L1 readers called morphological decomposition. While the practice of explicitly teaching morphological awareness skills is prevalent in ESL classes, more research is needed to establish what is known about gains in L2 morphological awareness, and its relationship to the development of automatic morphological decomposition processes in English language learners. The present study seeks to shed light on the nature of this relationship across growth in L2 proficiency. Two experimental measures were used: a masked priming paradigm with a lexical decision task to explore priming evidence for morphological decomposition and a paper and pencil test of morphological awareness which required subjects to derive the base of a morphologically complex word. These tasks were administered to L1 (N = 43) and L2 groups (intermediate N = 16, advanced N = 16) of university-aged subjects. Results indicated that all subjects show repetition priming effects. However, despite a significant gain in explicit knowledge of English morphology across proficiency levels, L2 learners don’t develop an ability to morphologically decompose words in the unconscious, automatic way that native English speakers do, as evidenced by a lack of morphological priming. Implications for L2 pedagogy and L2 word storage in the mental lexicon are discussed.