Line Integral Convolution Method Research Articles

Surface Line Integral Convolution-Based Vortex Detection Using Computer Vision

Abstract Vortex cores in fluid mechanics are easy to visualize, yet difficult to detect numerically. Precise knowledge of these allows fluid dynamics researchers to study complex flow structures and allow for a better understanding of the turbulence transition process and the development and evolution of flow instabilities, to name but a few relevant areas. Various approaches such as the Q, delta, and swirling strength criterion have been proposed to visualize vortical flows, and these approaches can be used to detect vortex core locations. Using these methods can result in spuriously detected vortex cores and which can be balanced by a cutoff filter, making these methods lack robustness. To overcome this shortcoming, we propose a new approach using convolutional neural networks to detect flow structures directly from streamline plots, using the line integral convolution method. We show that our computer vision-based approach is able to reduce the number of false positives and negatives while removing the need for a cutoff. We validate our approach using the Taylor–Green vortex problem to generate input images for our network. We show that with an increasing number of images used for training, we are able to monotonically reduce the number of false positives and negatives. We then apply our trained network to a different flow problem where vortices are still reliably detected. Thus, our study presents a robust approach that allows for reliable vortex detection which is applicable to a wide range of flow scenarios.

A Line Integral Convolution Method With Dynamically Determining Step Size and Interpolation Mode for Vector Field Visualization

The line integral convolution method can obtain a continuous and dense vector streamline texture. The traditional fixed step size gets larger texture noise, while the exact calculation step method (such as the Runge-Kutta method) is time-consuming. In the process of generating streamline texture, if using bilinear interpolation calculation, we can get the method of calculating the point vector value, and it is also time-consuming. To address the above-mentioned problems, the line integral convolution method for dynamically determining integral step size and interpolation mode is proposed. The effect of streamline drawing mainly depends on the size of the integration step and the interpolation mode. The method proposed in this paper gives the rules for determining the integration step size and interpolation mode according to the vector speed size and the angle between two vectors. The experimental results show that compared with the previous fixed step method and the bilinear interpolation method, the proposed method not only has faster computation speed but also clearer texture and stronger contrast.

Magnetic resonance angiography contrast enhancement and combined 3D visualization of cerebral vasculature and white matter pathways

Recently, in diffusion magnetic resonance imaging, the reconstruction and three-dimensional rendering of white matter pathways have been introduced to clinical routine protocols. In a number of clinical situations, for example the preoperative analysis of vascular pathologies, the assessment of spatial relations between vascular structures and nearby fiber pathways is of vital interest for treatment planning. In this paper, we present an approach to the integrated vessel and fiber visualization, based on a novel vascular contrast enhancement operator for Magnetic Resonance Angiography (MRA) datasets. We propose a 3D dynamic programming method, allowing contrast enhancement of vascular structures and suppression of partial voluming effects at vessel borders. This makes it easier to visualize vascular structures by realtime volume rendering with surface shading. In contrast to maximum intensity projection, the method provides better depth cues and allows for easier spatial orientation. The integration of tractography-generated fibers as streamlines or streamtubes with correct visibility computation is performed by a combined volume and geometry renderer. In situations where tractography fails to provide reliable results, we use a line integral convolution method to assess white matter structures. In this manner, the spatial relations of vessels to fiber structures can be depicted by three-dimensional visualizations. We evaluate our approach with clinical data from patients with arteriovenous malformations, stenoses, aneurysms, and from healthy volunteers.

Line integral convolution-based non-local structure tensor

The non-local structure tensors have received much attention recently. However, the current computation methods of non-local structure tensor fail to fully use the anisotropic characteristic of tensors, hence resulting in limited performance. To address this problem, we present a novel anisotropic non-local regularisation scheme that integrates the atomic decomposition strategy with an extended line integral convolution method using non-local means filtering technique, in order to sufficiently utilise the spatial direction relevancy of tensors for their anisotropic smoothing. Experimental results on the test images show that our proposed anisotropic non-local structure tensor is superior to the current representative nonlinear structure tensors in corner detection.

Influence of Geometry and Velocity of Rotating Solids on Hydrodynamics of a Confined Volume

Three cylinder-based geometries were evaluated at five different rotating speeds (ω = 20.94, 62.83, 94.25, 125.66, and 157.08 rad·s−1) to obtain the fluid flow pattern in nonsteady conditions. Two of the models were modified at the lower region, also known as tip section, by means of inverted and right truncated cone geometries, respectively. The experimental technique used a visualization cell and a Particle Imaging Velocimetry installation to obtain the vector field at the central plane of the volume. The Line Integral Convolution Method was used to obtain the fluid motion at the plane. In addition, the scalar kinetic energy and the time series were calculated to perform the normal probability plot. This procedure was used to determine the nonlinear fluid flow pattern. It was also used to identify two different flow regimens in physical and numerical results. As the rotation speed increased, the turbulent regions were placed together and moved. The process makes experimental observation difficult. The biphasic and turbulence constitutive equations were solved with the Computational Fluid Dynamics technique. Numerical results were compared with physical experiments for validation. The model with the inverted truncated cone tip presented better stability in the fluid flow pattern along the rotation speed range.

Line integral convolution based non-local structure tensor

The non-local structure tensors have received much attention recently. However, the current computation methods of non-local structure tensor fail to fully use the anisotropic characteristic of tensors, hence resulting in limited performance. To address this problem, we present a novel anisotropic non-local regularisation scheme that integrates the atomic decomposition strategy with an extended line integral convolution method using non-local means filtering technique, in order to sufficiently utilise the spatial direction relevancy of tensors for their anisotropic smoothing. Experimental results on the test images show that our proposed anisotropic non-local structure tensor is superior to the current representative nonlinear structure tensors in corner detection.

GPU accelerated marine data visualization method

The study of marine data visualization is of great value. Marine data, due to its large scale, random variation and multi-resolution in nature, are hard to be visualized and analyzed. Nowadays, constructing an ocean model and visualizing model results have become some of the most important research topics of ‘Digital Ocean’. In this paper, a spherical ray casting method is developed to improve the traditional ray-casting algorithm and to make efficient use of GPUs. Aiming at the ocean current data, a 3D view-dependent line integral convolution method is used, in which the spatial frequency is adapted according to the distance from a camera. The study is based on a 3D virtual reality and visualization engine, namely the VV-Ocean. Some interactive operations are also provided to highlight the interesting structures and the characteristics of volumetric data. Finally, the marine data gathered in the East China Sea are displayed and analyzed. The results show that the method meets the requirements of real-time and interactive rendering.

Influence of frame-dragging on magnetic null points near rotating black holes

Understanding the mechanisms of particle acceleration from the vicinity of black holes poses a challenge. Electromagnetic effects are thought to be a prime suspect, but details still need an explanation. To this end, we study a three-dimensional structure of oblique magnetic fields near a rotating black hole in vacuum. It has been proposed that such a setup can lead to efficient acceleration when plasma is injected near a magnetic null point. We focus our attention especially on the magnetic field in the immediate neighborhood of the magnetic null point, which was previously shown to occur in the equatorial plane. By employing the line integral convolution method, we visualize the magnetic field lines and explore the electric lines rising out of the equatorial plane. We show the magnetic field structure near the boundary of ergosphere, depending on the spin of the black hole. Electric field develops a non-vanishing component passing through the magnetic null point and ensuring efficient acceleration of charged particles from this particular location near the horizon. We also examine the effect of translatory boost on the field lines. Similarly to the frame-dragging by rotation, the linear motion carries field lines along with the black hole. The position of the magnetic null point recedes from the black hole horizon as the spin parameter increases. For the extreme value of a = 1, the null point can occur outside the ergosphere.

A Novel Method to Capture and Analyze Flow in a Gross Pollutant Trap Using Image-Based Vector Visualization

A novel method is developed to capture and analyze several experimental flow regimes through a gross pollutant trap (GPT) with fully and partially blocked screens. Typical flow conditions and screen blockages are based on findings from field investigations that show a high content of organic matter in urban areas. Fluid motion of neutral buoyant particles is tracked using a high-speed camera and particle image velocimetry (PIV) software. The recorded fluid motion is visualized through an image-based, line integral convolution (LIC) algorithm, generally suitable for large computational fluid dynamics (CFD) datasets. The LIC method, a dense representation of streamlines, is found to be superior to the point-based flow visualization (e.g., hedgehog or arrow plots) in highlighting main flow features that are important for understanding litter capture and retention in the GPT. Detailed comparisons are made between the flow regimes, and the results are compared with CFD data previously obtained for fully blocked screens. The LIC technique is a useful tool for identifying flow structures in the GPT and areas that are subjected to abnormalities difficult to detect by conventional methods. The novel method is found to be useful both in the laboratory and in the field, with little preparation and cost. The enhancements and pitfalls of the LIC technique along with the experimentally captured flow field are presented and discussed.

Output-Sensitive 3D Line Integral Convolution

We propose an output-sensitive visualization method for 3D line integral convolution (LIC) whose rendering speed is largely independent of the data set size and mostly governed by the complexity of the output on the image plane. Our approach of view-dependent visualization tightly links the LIC generation with the volume rendering of the LIC result in order to avoid the computation of unnecessary LIC points: early-ray termination and empty-space leaping techniques are used to skip the computation of the LIC integral in a lazy-evaluation approach; both ray casting and texture slicing can be used as volume-rendering techniques. The input noise is modeled in object space to allow for temporal coherence under object and camera motion. Different noise models are discussed, covering dense representations based on filtered white noise all the way to sparse representations similar to oriented LIC. Aliasing artifacts are avoided by frequency control over the 3D noise and by employing a 3D variant of MIPmapping. A range of illumination models is applied to the LIC streamlines: different codimension-2 lighting models and a novel gradient-based illumination model that relies on precomputed gradients and does not require any direct calculation of gradients after the LIC integral is evaluated. We discuss the issue of proper sampling of the LIC and volume-rendering integrals by employing a frequency-space analysis of the noise model and the precomputed gradients. Finally, we demonstrate that our visualization approach lends itself to a fast graphics processing unit (GPU) implementation that supports both steady and unsteady flow. Therefore, this 3D LIC method allows users to interactively explore 3D flow by means of high-quality, view-dependent, and adaptive LIC volume visualization. Applications to flow visualization in combination with feature extraction and focus-and-context visualization are described, a comparison to previous methods is provided, and a detailed performance analysis is included.

Visualization techniques in plasma numerical simulations

Numerical simulations of plasma processes usually yield a huge amount of raw numerical data. Information about electric and magnetic fields and particle positions and velocities can be typically obtained. There are two major ways of elaborating these data. First of them is calledplasma diagnostics. We can calculate average values, variances, correlations of variables, etc. These results may be directly comparable with experiments and serve as the typical quantitative output of plasma simulations. The second possibility is theplasma visualization. The results are qualitative only, but serve as vivid display of phenomena in the plasma followed-up. An experience with visualizing electric and magnetic fields via Line Integral Convolution method is described in the first part of the paper. The LIC method serves for visualization of vector fields in two dimensional section of the three dimensional plasma. The field values can be known only in grid points of three-dimensional grid. The second part of the paper is devoted to the visualization techniques of the charged particle motion. The colour tint can be used for particle’s temperature representation. The motion can be visualized by a trace fading away with the distance from the particle. In this manner the impressive animations of the particle motion can be achieved.

A visualizing application of line integral convolution techniques

The line integral convolution (LIC) technique, a texture synthesis technique, has served as a useful method for visualizing vector data. However, a number of shortcomings have been identified in the LIC technique, not the least of which are that the technique is computationally expensive and adopts a more or less brute force approach. This paper presents a modification of the original LIC method which addresses these shortcomings. Our method, although used for visualizing atmospheric vortical flows, is also applicable to other atmospheric phenomena. In our method we employ a particular colouring scheme to unambiguously identify the nature of the vortical flows irrespective of the hemisphere in which they occur.

Using ILIC algorithm for an impressionist effect and stylized virtual environments

Vector fields, often used in science, also find application in the artistic world of sculpture and painting. Apart from color and shape, artists utilize features of direction to depict our natural environment. The line integral convolution (LIC) method, originally developed for imaging vector field in scientific visualization, has the potential to produce images with directional characteristics. In this paper, we present four techniques to explore LIC in generating images in the style of the Impressionists. In particular, we develop an ILIC (Impressionist LIC) algorithm, which takes advantage of the directional information given by a photograph image, incorporates it with a shading technique to blend cool or warm colors into the image, and finally applies the LIC method to imitate artists’ paintings in the Impressionist style. The cool-to-warm blending scheme derives from a painting convention to enhance depth perception. Furthermore, we present a technique using statistical details to control the convolution length for the ILIC algorithm, allowing detailed information to be preserved during the convolution process. We also demonstrate a top-down sampling technique so that a series of mipmaps can be generated to construct virtual environments. These maps, even under multi-resolution viewing, reveal constant strokes, thereby achieving frame coherence in an interactive walkthrough system. The experimental results show satisfactory emulations and efficient computing in fabricating a wide category of artistic virtual environments.

A visualization of a vector-field by a homogenized nascent-particles tracking

Particle tracking methods have been developed to visualize a vector-field. The methods are efficient to show a large-scale and a small-scale structure in the flow field clearly. However, initial positions of the particles sometimes affect a visualized pattern. In addition, when a particle lies on another one, it becomes difficult to display the particle and to recognize the pattern. In this paper, we consider an enhancement of particle tracking methods against those problems. First, we attempt to reduce the dependence of pattern on the initial position and consider how to display a particle occupies another particle position. Essence of our solution is composed of homogenizing nascent-particles and a certain manipulation in a pixel/voxel space. This solution is similar to that of line integral convolution method (LIC). Secondly, accuracy of the visualization process is discussed. Then, we make clear some pitfalls of LIC. Finally, a new technique to overcome drawbacks of conventional particle tracking methods and LIC is described. Several flow fields in two-dimensions and near surfaces are visualized for demonstration of our proposed method.