Knuth,M.,Kohlhammer,J.
Abstract:
In this paper we present a hybrid technique for illuminating a dynamic scene with self shadowing. Our goal is to perform the necessary calculations with real-time or interactive frame rates. The main idea is to split up the self shadowing process into a global and a local part. This separation allows us to choose combinations of completely independent algorithmic approaches. The global part calculates the light occlusion of the entire scene. Since it has to process the whole scene, it has a high computational cost, but a coarse approximation of the global occlusion can be created in a short time. On the other hand the local part deals with fine details which have local impact to the scene. It only processes a relevant subset of the scene. In our case this subset is given by the part of the scene, which is currently visible on the screen. Our idea is to use a GPU based approach for this part, which completely works inside image space and is independent from the scene complexity.
Vergeest,J.S.M., Song,Y.
Abstract:
We present an approach of geometric fitting of freeform shapes to 3D data points, where the size of the fitting problem is relatively large. The shapes studied represent ship propeller blades of dimension up to 3m, with precision requirements of few mm. Moreover, the shapes are freeform and designed using a geometric model dependent on hundreds of numerical parameters. For the ship industry it is crucial to optimally deal with the shape parameters in order to judge whether or not a particular manufactured part fits within the tolerances. We have developed a method to evaluate the shapes numerically and we report on the approach we took. The sensitivity of a deviation objective function with respect to critical design parameters could be acquired. Also some procedures for automated optimization were explored.
Ren,G., Yang,X.
Abstract:
A novel blind robust 3D mesh object watermarking method is proposed. Firstly, we choose the prongs of the mesh; secondly, to suppose that the prongs are the centers of the circle and the length we choose is the radius, we can get the areas which could be watermarked; and then, with the sphere coordinate, we will get the distance from the object center to the vertices, and perform DFT on the sphere radiuses; the watermark will be embedded by changing the DFT frequency coefficients. Experimental results show that this watermarking scheme is robust against the attacks such as rotation, translation, uniform scaling and even cropping and the embedded watermark is invisible.
Han,D., Yang,X.
Abstract:
The main objective of 3D mesh watermarking is to maximize the watermark strength while minimizing the visual distortion. In this paper, we propose a novel 3D mesh watermarking based on the human visual system and mesh curvature. We analyze the region’s mean curvature and fluctuation of curvature, and then choose the appropriate regions with high visual masking degree, in order to embed the watermark by using the Interpolation Algorithm. The experimental results show that the watermark is capacious, invisible and robust against various attacks.
Yasmin,S., Talib,A.Z.
Abstract:
3D shape transformation is usually confined to transformation between a pair of objects. The objective of this paper is to look at shape transformation from a different perspective: instead of binding this concept between two objects, the technique is extended to the concept of incorporating the characteristics of a number of objects in one body at a time. Equal number of slices are generated from all objects. Slices may be parallel to each other or each slice may have different orientation. Traversal of a data along its longitudinal direction may generate slices which are differently oriented from each other. When multiple objects are transformed to one and is used as an influence shape, it also works as incorporating multiple influence shapes at a time during transformation between two objects. The paper shows the ease of implementation of this concept in sliced data and also discusses its extendibility.
Sisojevs,A., Krechetova,K., Glazs,A.
Abstract:
This work describes an efficient method of 3D object modeling. The method is applied for solving biomedical engineering tasks. The method is based on interpolation model creation. The model is built using Coons surface basis. For further implementation of the model the Coons surface basis is transformed to Bezier surface model. This work also describes a method of volume estimation of the reconstructed 3D model. Example of practical implementation is solving various biomedical tasks.
Cabral,I.P.S., Gonçalves,L.M.G., XavierJ.C.Jr.
Abstract:
The use of Geographic Information Systems (GIS) has becoming very important in fields where detailed and precise study of earth surface features is required. Environmental protection is such an example that requires the use of GIS tools for analysis and decision by managers and enrolled community of protected areas. In this specific field, a challenge that remains is to build a GIS that can be dynamically fed with data, allowing researchers and other agents to recover actual and up to date information. In some cases, data is acquired in several ways and come from different sources. To solve this problem, we propose a tool that includes a model for spatial data treatment on the Web, that we named System for Integrated Monitoring - SIM. The research issues involved in this tool start with the dealing of satellite images of protected areas. It is followed by acquisition and processing of different types of video and images as Small Format Aerial Images, or simply SFAI, acquired by a radio controlled helicopter, used for construction of mosaics to be fed in the GIS, and underwater geo-referenced pictures, acquired in-loco by researchers for better analyzing a given area. The SIM continues with the feeding and processing of environmental control data collected in-loco as biotic and geological variables and finishes with the presentation of all information on the Web. For this dynamic processing, we have developed tools that make MapServer more flexible and dynamic, allowing data uploading by the proper users. As example, on the top of SIM, we have developed a module that uses interpolation to aiming spatial data analysis. A very complex application that has validated our research is to feed the system with data coming from coral reef regions located in northeast of Brazil.
Han,E., Kim,N., Hong,H.
Abstract:
The acceleration structures of the scene such as BVH(Bounding Volume Hierarchy) and KD-tree have been widely used to improve the rendering performance. This paper presents a novel rendering algorithm based on spherical coordinate representation of the object. The vertices of the object are transformed into spherical coordinate system, and we construct three maps: the centroid and index of the triangle, the memory access table. While OpenGL pipeline touches all vertices of the object, the proposed method takes account the only visible vertices by efficiently examining the visible triangles in spherical representation. Simulation results demonstrated that the proposed method achieves an efficient rendering performance.
Lee,S.-H., Han,S.-I., Kang,M.G.
Abstract:
We propose a motion detection algorithm which works well in low illumination environments. By using the level set based bimodal motion segmentation, the algorithm obtains an automatic segmentation of the motion region and the spurious regions due to
the large CCD noise in low illumination environment are removed effectively. Based on a noise analysis, we will show how to obtain the required parameters automatically. Experimental results verify the stableness of the proposed algorithm in low illumination conditions.
Puig-Centelles,A., Varley,P.A.C., Ripolles,O., Chover,M.
Abstract:
Virtual environments should offer the user a deep interactive experience with both large worlds to explore and a higher degree of perceived realism. The main goal of our work is to provide the final user with an easy-to-use accurate terrain generation application, which allows non-professional users to design their own desired terrain. In this paper we consider the creation of islands to be used in computer games. We introduce a simple terrain algorithm and we also consider its integration into a sketching application. The application will offer both a 2D and a 3D representation of the terrain, in order to simplify the interface and provide the user with more interactive feedback about the island that has been designed. Our framework offers real-time algorithms for both creating and modifying terrain features, thus improving the final results with more realism and greater customization by the user.
Löffler,F., Rybacki,S., Schumann,H.
Abstract:
The interactive visualisation of digital terrain datasets deals with there interrelated issues: quality, time and resources. In this
paper a GPU-supported rendering technique is introduced, which finds a tradeoff between these issues. For this we use the
projective grid method as the foundation. Even though the method is simple and powerful, its most significant problem is
the loss of relevant features. Our contribution is a definition of a view-dependent grid distribution on the view-plane and an
error-bounded rendering. This leads to a better approximation of the original terrain surface compared to previous GPU-based
approaches. A higher quality is achieved with respect to the grid resolution. Furthermore the combination with an error metric
and ray casting enables us to render a terrain representation within a given error threshold. Hence, high quality interactive
terrain rendering is guaranteed, without expensive preprocessing.
Aguiar,C.R.S., Druon,S., Crosnier,A.
Abstract:
In this paper, we present a new algorithm for the alignment of two 3D scans. The approach uses a region-based matching technique. Regions are described by a probability density function (pdf) computed from low dimensional surface descriptors (curvature or normal cone). Region correspondence is found using similarity function based on the comparison of regions pdf and under geometry constraints. Results on raw scan data sets are presented to illustrate the algorithm.
Kim,K.R., Lee,B.G., Yoo,H.
Abstract:
This paper presents an improved method for the direction-oriented interpolation (DOI) method. The technique of DOI is considered to be a very strong tool for intraframe-based deinterlacing in the literature. However, the DOI still have a problem to use wrong edge direction. To remedy this problem, we embed three steps in the DOI method to interpolate missing lines more efficiently and robustly than existing methods. In the proposed method, the spatial direction vector (SDV) data are reused and processed to prevent them utilizing in next interpolation step, resulting in the more accurate de-interlacing method. We carry out experiments to compare the proposed method with the existing methods including the edge-based line averaging (ELA) and DOI. Experimental results show that the proposed method gives better performance in objective and subjective quality than the existing de-interlacing methods.
Zemčík,P., Maršík,L., Herout,A.
Abstract:
This contribution describes recent development in ongoing work focused on point cloud rendering algorithm implementation usable in environments containing programmable or custom hardware. The approach described in this paper is based on the idea that direct point cloud rendering, which is in the principle not too complicated, can be efficiently implemented in programmable or custom hardware. Such implementation can be useful not only for its performance but especially for the possibility to include it into solutions that require 3D graphics output in non PC environments and in embedded solutions with low power consumption, etc. This contribution describes the overall approach and the current results.
Mlích,J., Zemčík,P., Jiřík,L.
Abstract:
This paper focuses on evaluation of motion of objects through classification of their trajectories. The objects used for evaluation of the presented method are objects tracked in video sequence but the method can be used for more general trajectories. The main potential application of the presented method is detection of abnormal behavior of humans, hand gesture detection and recognition, etc. Hidden Markov Models (HMMs) are used as the classification mechanism. The paper contains description of the method, description of experiments and their results, and conclusions.
Haindl,M. Hatka,M.
Abstract:
This paper describes a generalization of our previously published simple roller method for seamless enlargement of colour textures such as natural bidirectional texture functions (BTF) that realistically represent appearance of given material surfaces. The generalized roller allows automatic detection of major texture periodicity directions which do not need to be aligned with coordinate axes. The roller exture synthesis method is based on the overlapping tiling and subsequent minimum error boundary cut. One or several optimal double toroidal BTF patches are seamlessly repeated during the synthesis step. While the method allows only moderate texture compression it is extremely fast due to complete separation of the analytical step of the algorithm from the texture synthesis part. The method is universal and easily implementable in a graphical hardware for pe of static or dynamic textures.
Yang,X., Xu,D.-Q.,Zhao,L.
Abstract:
Recent GPU ray tracers can already achieve performance competitive to that of their CPU counterparts. Nevertheless, these systems can not yet fully exploit the capabilities of modern GPUs and can only handle medium-sized, static scenes.
We present an octree construction algorithm for the GPU that achieves real-time performance by heavily exploiting the hardware, which has been observed to give superior performance in ray tracing compared to other acceleration structures. We use streaming construction with the surface area heuristic (SAH) that significantly increase the coherence of memory accesses during construction of the octree.
Lefer,W.
Abstract:
Crowd simulation has received increasing attention for two decades because potential applications of crowd simulators can be found in various societal domains. The continuum crowd model allows to integrate all information useful for the decision-making process in a single equation, which can then be solved by a Fast Marching Method approach. In this paper we propose several improvements to the continuum crowd model: a new governing equation, a new collision avoidance method, and, our major contribution, we add vision capabilities to the characters, thus making them able to collect new information about their surrounding environment and to reconsider path planning according to up-to-date data.
Beran,V., Herout,A., Řezníček,I.
Abstract:
Bicycle detection approach covered by this paper aims to cope with highly-noisy low-resolution data, to use simple image-processing methods and to work in real time. Although the method itself does not constitute a generally usable object detector, it covers several interesting aspects which can be re-used in tasks similar to the given one. Low-level features extracted from the video used for wheel-candidate classification are described in detail. The system is applied and evaluated on real data and the results are discussed.
Garnier,L., Belbis,B., Foufou,S.
Abstract:
Toruses and double spheres are particular cases of Dupin cyclides. In this paper, we study the conversion of rational biquadratic Bézier surfaces into Dupin cyclide patches.
We give the conditions that the Bézier surface should satisfy to be convertible, and present a new conversion algorithm to construct the torus or double sphere patch corresponding to a given Bézier surface, some conversion examples are illustrated and commented.
Wibirama,W., Pintavirooj,Ch., Tungjitkusolmun,S., Hamamoto,K.
Abstract:
Robust, low cost and versatile eye tracking system has been a fundamental and challenging problem for biomedical engineering and computer vision area. Currently, almost all eye tracking researchs are done by video oculography method which is based on several image processing tasks to estimate the trajectory of eye gaze. This paper reports a new method to estimate eye-motion position and direction based on improved window-matching and gradient analysis technique. The technique is implemented in low cost eye tracking system consists of two infrared cameras attached to special binocular. Experimental results show that the method improves the accuracy of tracking pupil position in large and less head movement condition.
Thourn,K., Kitjaidure,Y.
Abstract:
This paper presents the combination of Principal Component Analysis (PCA) and Fisher Linear Discriminant Analysis (FLDA) for affine invariant shape recognition. This technique extracts the 2D shapes feature using the signed enclosed area from decomposed images in different scale levels of Barycenter contour. These shapes features are represented in 3D format. After computing the shape representations, they are passed through PCA for dimensionality reduction. However PCA technique is not invariant to the starting point, so the Discrete Fourier Transform (DFT) is applied to the shapes features. Next FLDA is used for making discrimination. Finally, the similarity is measured by the Euclidian distance. The experimentation is evaluated with affine invariant shape dataset of 560 shapes divided into 40 different classes and 14 affine distorted shapes in each class. The experiment results show that this approach gives high recognition rate.
Schoor,W., Seidl,T., Bollenbeck,F., Seiffert,U., Preim,B., Mecke,R.
Abstract:
This paper presents a novel method for efficient semiautomatic multi-label segmentation of plant biomedical image data. The approach extends live-wire methods in order to facilitate exact user-steered segmentations for atlas generation. By integrating a segmentation-specific user interaction model into the live-wire formulation i) more exact segmentations, ii) increased computational efficiency, iii) without loss of generality are achieved. The concept of mutual influence of image feature based path costs and user input uncertainty are consistently combined. By incorporating user behavior into cost based delineation a more intuitive user interface is obtained also yielding in a more accurate segmentation. We introduce path-based methodologies, specific user interaction models and propose the combination of both of them. The purposefulness of the method is shown in an application comprising segmentation of histological section data supporting the generation of the 3D atlases.
Benger,W., Ritter,M., Acharya,S., Roy,S., Jijao,F.
Abstract:
We describe a novel approach to treat data from a complex numerical simulation in a unified environment using a generic data model for scientific visualization. The model is constructed out of building blocks in a hierarchical scheme of seven levels, out of which only three are exposed to the end-user. This generic scheme allows for a wide variety of input formats, and results in powerful capabilities to connect data. We review the theory of this data model, implementation aspects in our visualization environment, and its application to computational fluid dynamic simulation covering a fluid in a stir tank. The computational data are given as a vector field and a scalar field describing pressure on 2088 blocks in curvilinear coordinates