VMV11

Permanent URI for this collection

https://diglib.eg.org/handle/10.2312/1015

Browse

Now showing 1 - 20 of 48

A Mathematical Model and Calibration Procedure for Galvanometric Laser Scanning Systems
(The Eurographics Association, 2011) Manakov, Alkhazur; Seidel, Hans-Peter; Ihrke, Ivo; Peter Eisert and Joachim Hornegger and Konrad Polthier
Laser galvanometric scanning systems are commonly used in various fields such as three dimensional scanning, medical imaging, material processing, measurement devices and laser display systems. The systems of such kind suffer from distortions. On top of that they do not have a center of projection, which makes it impossible to use common projector calibration procedures. The paper presents a novel mathematical model to predict the image distortions caused by galvanometric mirror scanning systems. In addition, we describe a calibration procedure for recovering its intrinsic and extrinsic parameters.
SBL Mesh Filter: A Fast Separable Approximation of Bilateral Mesh Filtering
(The Eurographics Association, 2011) Vialaneix, Guillaume; Boubekeur, Tamy; Peter Eisert and Joachim Hornegger and Konrad Polthier
Bilateral mesh filtering is a simple and powerful feature-preserving filtering operator which allows to smooth or remove noise from surface meshes while preserving important features in a non-iterative way. However, to be effective, such a filter may require quite a large support size, inducing slow processing when applied on high resolution meshes such as the ones produced by automatic 3D acquisition devices. In this paper, we propose a separable approximation of bilateral mesh filtering based on a local decomposition of the bi-dimensional filter into a product of two one-dimensional ones. In particular, we show that this approximation leads to piecewise smooth surfaces which are very close to the ones produced by the exact filter, using only a fraction of the usual required time. Compared to bilateral image filtering, the main problem here is to find meaningful directions at every point to orient the two one-dimensional filters. Our solution exploits the minimum and maximum curvature directions at each point and demonstrates a significant speed-up on meshes ranging from thousands to millions of elements, enabling feature-preserving filtering with large support size in a variety of practical scenarii. Our approach is simple, easy to implement and orthogonal to other kinds of optimizations such as higher dimensional clustering using a bilateral grid or a Gaussian kd-tree and can therefore be combined to them to reach even higher performance.
Simulating Deep Sea Underwater Images Using Physical Models for Light Attenuation, Scattering, and Refraction
(The Eurographics Association, 2011) Sedlazeck, Anne; Koch, Reinhard; Peter Eisert and Joachim Hornegger and Konrad Polthier
When adapting computer vision algorithms to underwater imaging, two major differences in image formation occur. While still traveling through the water, light rays are scattered and absorbed depending on their wavelength, creating the typical blue hue and low contrast in underwater images. When entering the underwater housing of the camera, light rays are refracted twice upon passing from water into glass and into air. We propose a simulator for both effects based on physical models for deep sea underwater images captured by cameras in underwater housings with glass port thicknesses in the order of centimeters. Hence, modeling refraction by explicitly computing the correct path of the rays allows to accurately simulate distortions induced by underwater housings. The Jaffe- McGlamery model for effects on color is often used in computer vision algorithms as a base for simplification. We extend this model to incorporate color images, shadows, and several light sources.
Seamless Integration of Multimodal Shader Compositing into a Flexible Ray Casting Pipeline
(The Eurographics Association, 2011) Arens, Stephan; Bolte, Matthias; Domik, Gitta; Peter Eisert and Joachim Hornegger and Konrad Polthier
In the last three years a number of multi-volume GPU ray casting systems have been presented. Some of them are very powerful and provide a wide variety of features. However, these approaches are either only capable of displaying multiple modalities together without logically combining them or they lack the necessary flexibility for rapid visual development. These features are fundamental for visualizing the coherent information of multimodal data. In this paper we therefore present an integrated way of visually specifying a volume rendering pipeline including a flexible multimodal compositing of sampling, transfer functions, logical operators and shading. As a result the data flow can be visually constructed and retraced from preprocessing through to the shader operations. Hence intuitive visual prototyping of multimodal transfer function compositing is possible at runtime.
A Global Optimization Approach to High-detail Reconstruction of the Head
(The Eurographics Association, 2011) Schneider, David C.; Kettern, Markus; Hilsmann, Anna; Eisert, Peter; Peter Eisert and Joachim Hornegger and Konrad Polthier
The paper presents an approach for reconstructing head-and-shoulder portraits of people from calibrated stereo images with a high level of geometric detail. In contrast to many existing systems, our reconstructions cover the full head, including hair. This is achieved using a global intensity-based optimization approach which is stated as a parametric warp estimation problem and solved in a robust Gauss-Newton framework. We formulate a computationally efficient warp function for mesh-based estimation of depth which is based on a well known image-registration approach and adapted to the problem of 3D reconstruction. We address the use of sparse correspondence estimates for initializing the optimization as well as a coarse-to-fine scheme for reconstructing without specific initialization. We discuss issues of regularization and brightness constancy violations and show various results to demonstrate the effectiveness of the approach.
AVDT - Automatic Visualization of Descriptive Texts
(The Eurographics Association, 2011) Spika, Christian; Schwarz, Katharina; Dammertz, Holger; Lensch, Hendrik P. A.; Peter Eisert and Joachim Hornegger and Konrad Polthier
Expressing mental images visually as 3D scenes is a time-consuming challenge. Therefore, we employ natural language to facilitate the creation of virtual environments. In this paper, we present a framework, which automatically converts an arbitrary descriptive text into a representative 3D scene. Our system parses a user-written input text, extracts information using techniques from Natural Language Processing (NLP) and identifies relevant units. Based on derived object-to-object relations, our system associates every object with an appropriate 3D model and evaluates spatial dependencies of the entities. The resulting locations are combined based on adequate heuristics in order to create natural looking virtual environments. Finally, a physics engine is used to render a realistic and interactive 3D scene which enables the user to actively manipulate the stage setup.
Object-aware Gradient-Domain Image Compositing
(The Eurographics Association, 2011) Eisemann, Martin; Kokemüller, Jan; Magnor, Marcus; Peter Eisert and Joachim Hornegger and Konrad Polthier
We describe an approach to suppress bleeding artifacts without altering the boundary location in gradient-domain compositing, a technique to create seamless composites. While gradient-domain compositing has become a stan- dard tool for many complex image editing tasks such as seamless cloning, panorama stitching or scene completion, its quality suffers from mismatches in the composited image regions. We propose an approach that is robust to non- optimal region selection by the user without altering his selection which may be neither intended nor possible for certain compositing tasks. In addition, we present an easy-to-use extension to composite interleaving objects. The usability of our approach is demonstrated by several image compositing tasks and comparisons to current state-of-the-art algorithms in gradient-domain image compositing are presented.
Smooth Transitions for Large Scale Changes in Multi-Resolution Images
(The Eurographics Association, 2011) Lancelle, Marcel; Fellner, Dieter W.; Peter Eisert and Joachim Hornegger and Konrad Polthier
Today's super zoom cameras offer a large optical zoom range of over 30x. It is easy to take a wide angle photograph of the scene together with a few zoomed in high resolution crops. Only little work has been done to appropriately display the high resolution photo as an inset. Usually, to hide the resolution transition, alpha blending is used. Visible transition boundaries or ghosting artifacts may result. In this paper we introduce a different, novel approach to overcome these problems. Across the transition, we gradually attenuate the maximum image frequency. We achieve this with a Gaussian blur with an exponentially increasing standard deviation.
A Framework for Interactive Visualization and Classification of Dynamical Processes at the Water Surface
(The Eurographics Association, 2011) Wanner, Sven; Sommer, Christoph; Rocholz, Roland; Jung, Michael; Hamprecht, Fred; Jähne, Bernd; Peter Eisert and Joachim Hornegger and Konrad Polthier
A framework for the visualization and classification of multi-channel spatio-temporal data from water wave imaging is presented. Our interactive visualization tool, WaveVis, allows a detailed study of the water surface shape in reference to additional data streams, like thermographic images or classification results. This facilitates an intuitive and effective inspection of huge amounts of data. WaveVis was used to select representative training examples of events for a supervised learning approach and to evaluate the results of the classification. The interactive classification and segmentation software ilastik was used to train a Random Forest classifier. The benefit of the combination of both programs is demonstrated for two applications, the estimation of the rain rate from the segmentation of impact craters, and the detection of small scale breaking waves. The classification of the impact crater of raindrops on the water surface worked very well, whereas the detection of the breaking waves was satisfactory only under certain experimental conditions. Nevertheless, the combination of WaveVis and ilastik proved to be valuable in both cases.
Surface Flow from Visual Cues
(The Eurographics Association, 2011) Petit, Benjamin; Letouzey, Antoine; Boyer, Edmond; Franco, Jean-Sébastien; Peter Eisert and Joachim Hornegger and Konrad Polthier
In this paper we study the estimation of dense, instantaneous 3D motion fields over a non-rigidly moving surface observed by multi-camera systems. The motivation arises from multi-camera applications that require motion information, for arbitrary subjects, in order to perform tasks such as surface tracking or segmentation. To this aim, we present a novel framework that allows to efficiently compute dense 3D displacement fields using low level visual cues and geometric constraints. The main contribution is a unified framework that combines flow constraints for small displacements with temporal feature constraints for large displacements and fuses them over the surface using local rigidity constraints. The resulting linear optimization problem allows for variational solutions and fast implementations. Experiments conducted on synthetic and real data demonstrate the respective roles of flow and feature constraints as well as their ability to provide robust surface motion cues when combined.
A Cluster Hierarchy-based Volume Rendering Approach for Interactive Visual Exploration of Multi-variate Volume Data
(The Eurographics Association, 2011) Dobrev, Petar; Long, Tran Van; Linsen, Lars; Peter Eisert and Joachim Hornegger and Konrad Polthier
Interactive visual analysis of volumetric data relies on intuitive, yet powerful mechanisms to generate transfer functions. For multi-variate data, traditional methods for interactive transfer functions generation are of limited use. We propose a novel approach, where the user operates in a cluster space. It relies on hierarchical densitybased clustering of the high-dimensional feature space. The cluster tree visualization in form of a 2D radial layout serves as an interaction widget for selecting clusters, assigning material properties, changing sizes, merging, and splitting. This widget is complemented by a linked parallel coordinates widget. The interactive selections are automatically mapped to a transfer function for a linked 3D texture-based direct volume rendering, where brushing in parallel coordinates leads to the generation of a 3D binary opacity mask that is overlaid with the opacity values obtained from cluster tree selections. In GPU memory, we only hold the density values from the clustering approach and the cluster IDs. The derived density field allows us to interactively change the size of clusters and to compute normals for lighting. We applied our methods to the visualization of multi-variate data consisting of multiple scalar fields as well as derived scalar property fields from single scalar and vector fields. Our approach scales well to arbitrarily high dimensionality as the complexity of the main user interactions do not increase with the number of dimensions.
Adaptive Sampling for Geometry-aware Reconstruction Filters
(The Eurographics Association, 2011) Bauszat, Pablo; Eisemann, Martin; Magnor, Marcus; Peter Eisert and Joachim Hornegger and Konrad Polthier
We present an adaptive sampling scheme for Monte-Carlo-based renderers with the aim to support geometryaware filtering techniques for interactive computation of global illumination. While sophisticated filtering techniques for homogeneous areas can already produce high-quality results with as few as one sample per pixel, these approaches lack the ability to filter sufficiently in the vicinity of complex geometric structures. The result are visible artifacts in the final rendering result. Our sampling scheme distributes the samples for the indirect illumination in the image plane according to the necessity of a geometry-aware filtering. We show how to implement our scheme efficiently on current graphics hardware and how to combine it with a sophisticated filtering in order to achieve high-quality interactive frame rates for global illumination simulations. The resulting computational overhead is only in the range of a few milliseconds, making our approach suitable for real-time implementations.
Markerless Motion Capture using multiple Color-Depth Sensors
(The Eurographics Association, 2011) Berger, Kai; Ruhl, Kai; Schroeder, Yannic; Bruemmer, Christian; Scholz, Alexander; Magnor, Marcus; Peter Eisert and Joachim Hornegger and Konrad Polthier
With the advent of the Microsoft Kinect, renewed focus has been put on monocular depth-based motion capturing. However, this approach is limited in that an actor has to move facing the camera. Due to the active light nature of the sensor, no more than one device has been used for motion capturing so far. In effect, any pose estimation must fail for poses occluded to the depth camera. Our work investigates on reducing or mitigating the detrimental effects of multiple active light emitters, thereby allowing motion capture from all angles. We systematically evaluate the concurrent use of one to four Kinects, including calibration, error measures and analysis, and present a time-multiplexing approach.
Warp-based Near-Regular Texture Analysis for Image-based Texture Overlay
(The Eurographics Association, 2011) Hilsmann, Anna; Schneider, David C.; Eisert, Peter; Peter Eisert and Joachim Hornegger and Konrad Polthier
Image-based texture overlay or retexturing is the process of augmenting a surface in an image or a video sequence with a new, synthetic texture. Some properties of the original texture such as texture distortion as well as lighting conditions should be preserved for a realistic appearance of the augmented result. One approach would be to estimate a 3-dimensional geometry of the surface. However, this is an ill-posed problem for complex deformed surfaces like cloth, especially if only one image is given. In an image-based approach, these properties are directly estimated from the image. The key challenge is to separate the shading information from the actual local texture and to retrieve the texture distortion from an image without any knowledge of the underlying scene. In this paper, we model an image of a deformed regular texture as a combination of its deformed surface albedo, a shading map and additional high frequency details. We present a method for determination of these intrinsic parts of a given texture image by first estimating the appearance of a small texture element and then synthesizing a reference image of the undeformed regular texture. In a subsequent image-based optimization method this reference image is iteratively warped spatially and photometrically onto the original image whilst estimating deformation and illumination parameters. The decomposition is used to create images of new textures with the same deformation and illumination properties as in the original image.
The Recognition of Ethnic Groups based on Histological Skin Properties
(The Eurographics Association, 2011) Malskies, Christoph R.; Eibenberger, Eva; Angelopoulou, Elli; Peter Eisert and Joachim Hornegger and Konrad Polthier
We present an algorithm to recognize ethnic groups based on biologically justified features, such as melanin or hemoglobin concentrations. These biophysical features are extracted from skin reflectance spectra and allow, in contrast to technical features, a medical interpretation and intuitive rating of the recognition results. For this purpose, a physics-based light transport model for skin is required. We use an existing model based on Kubelka-Munk theory, which is physically accurate and computationally tractable. The evaluation of the ethnicity classification reveals that in comparison to an approach, directly based on the reflectance spectra, our proposed biophysical classification is slightly better. To reduce computation time we analyze the impact of spectral band reduction on the ethnicity classification and show that this can be achieved on the expense of only a small accuracy loss.
FreeCam: A Hybrid Camera System for Interactive Free-Viewpoint Video
(The Eurographics Association, 2011) Kuster, Claudia; Popa, Tiberiu; Zach, Christopher; Gotsman, Craig; Gross, Markus; Peter Eisert and Joachim Hornegger and Konrad Polthier
We describe FreeCam - a system capable of generating live free-viewpoint video by simulating the output of a virtual camera moving through a dynamic scene. The FreeCam sensing hardware consists of a small number of static color video cameras and state-of-the-art Kinect depth sensors, and the FreeCam software uses a number of advanced GPU processing and rendering techniques to seamlessly merge the input streams, providing a pleasant user experience. A system such as FreeCam is critical for applications such as telepresence, 3D video-conferencing and interactive 3D TV. FreeCam may also be used to produce multi-view video, which is critical to drive newgeneration autostereoscopic lenticular 3D displays.
Interactive Exploration of Polymer-Solvent Interactions
(The Eurographics Association, 2011) Thomaß, Bertram; Walter, Jonathan; Krone, Michael; Hasse, Hans; Ertl, Thomas; Peter Eisert and Joachim Hornegger and Konrad Polthier
The interaction of three-dimensional linked hydrophilic polymers with surrounding solvents in time-dependent data sets is of great interest for domain experts and current research in molecular dynamics. These polymers are called hydrogels, and their most characteristic property is their swelling in aqueous solutions by absorbing the solvent. Their conformation transition can be studied by investigations of the interaction of the single polymer strand and the solvent directly around the polymer at an atomistic level. We present new visualization techniques to interactively study time-dependent data sets from molecular dynamics simulations-with special regard to polymer-solvent interactions like local concentrations and hydrogen bonds-as well as filtering methods to facilitate analysis. Such methods that visualize polymer-solvent interactions on a hydration shell around a polymer are not available in current tools and can greatly facilitate the visual analysis, which helps domain experts to extract additional information about hydrogel characteristics and gain new insights from the simulation results. While our visual analysis methods presented in this paper clearly facilitate the analysis of hydrogels and lead to new insight, the presented concepts are applicable to other domains like proteins or polymers in general that interact with solvents.
Effective Back-Patch Culling for Hardware Tessellation
(The Eurographics Association, 2011) Loop, Charles; Nießner, Matthias; Eisenacher, Christian; Peter Eisert and Joachim Hornegger and Konrad Polthier
When rendering objects with hardware tessellation, back-facing patches should be culled as early as possible to avoid unnecessary surface evaluations, and setup costs for the tessellator and rasterizer. For dynamic objects the popular cone-of-normals approach is usually approximated using tangent and bitangent cones. This is faster to compute, but less effective. We present a novel approach using the Bézier convex hull of the parametric tangent plane. It is much more accurate, and by operating in clip space we are able to reduce the computational cost significantly. As our algorithm vectorizes well, we observe comparable test times with increased cull-rates.
Data-Driven Visualization of Functional Brain Regions from Resting State fMRI Data
(The Eurographics Association, 2011) Crippa, Alessandro; Roerdink, Jos B.T.M.; Peter Eisert and Joachim Hornegger and Konrad Polthier
Functional parcellation of the human cortex plays an important role in the understanding of brain functions. Tradi- tionally, functional areas are defined according to anatomical landmarks. Recently, new techniques were proposed that do not require a priori segmentation of the cortex. Such methods allow functional parcellation by functional information alone. We propose here a data-driven approach for the exploration of functional connectivity of the cortex. The method extends a known parcellation method, used in multichannel EEG analysis, to define and extract functional units (FUs), i.e., spatially connected brain regions that record highly correlated fMRI signals. We apply the method to the study of fMRI data and provide a visualization, inspired by the EEG case, that uses linked views to facilitate the understanding of both the location and the functional similarity of brain regions. Initial feedback on our approach was received from four domain experts, researchers in the field of neuroscience.
Fast and Efficient 3D Chamfer Distance Transform for Polygonal Meshes
(The Eurographics Association, 2011) Martinek, Michael; Grosso, Roberto; Greiner, Günther; Peter Eisert and Joachim Hornegger and Konrad Polthier
We present an efficient GPU-based method to perform 3D chamfer distance transform (CDT) in a wavefront scheme. In this context, we also introduce a binary voxelization algorithm which provides the initial boundary condition for the CDT. The voxelization method is capable of both, surface and solid voxelization, allowing for the computation of unsigned distance fields for arbitrary polygonal meshes and signed distances for models with orientable surfaces. Our method is trimmed on speed rather than accuracy. It works with simple chamfer metrics such as the Manhattan and chessboard distance and requires only a single rendering pass per distance layer. Due to the wavefront scheme, a propagation can be stopped if a required number of distance layers is reached. However, even a complete distance field can be computed in the magnitude of 10?3 seconds including the preprocessing voxelization step. This allows for a use in real-time applications such as path planning or proximity computations. We demonstrate the application of our method for the latter.

Browse

Browsing VMV11 by Issue Date

Results Per Page

Sort Options