Kotas,P.; Praks,P.; Valek,L.
Abstract:
In this paper an alternative method for the automatic pattern classification of metallographic images is presented. The aim of the pattern classification is to help monitoring the process quality in the steel plant of the company ArcellorMittal Ostrava plc, Ostrava, the Czech Republic. The here reprented approach is based on the well known Gabor filter, which provides suitable results in various texture analysis applications. In our case, the real metallographic samples are firstly separated from the image background. Then, a texture extraction is provided. The extracted samples are processed by applying the Gabor filter with various properties, from which selected texture features are formed. Effects of a dimension reduction technique for quality of similarity retrieval are studied.
Diaz-Tula,A., Castaneda-Garay,M., Belmonte-Fernandez,O.
Abstract:
The human-computer interaction using large projection screens is gaining more space nowadays. For these screens several computer vision techniques have been developed that allow the user to interact with the system through the projected images using laser pointers, special pens and the hands. On this work is presented the parallelization of a method for the real-time detection of non-stationary photometric perturbations in projection screens using the Computed Unified Device Architecture, in order to overcome the elevated running time of the serialized implementation on CPU. A comparison of the results is presented to establish the acceleration of the parallel algorithm against its original version on CPU.
Havel,J., Herout,A.
Abstract:
This paper describes some concepts from category theory, which are commonly used in functional programming. These concepts are applied to shader programming and to the rendering pipeline and the whole rendering pipeline is formally modelled using category theory. This model can be used for more abstract and formal approach to shader programming. Mathematical formalization of the rendering pipeline and its stages can be helpful in shader compiler design, for proving algorithms, complexity analysis, and other tasks.
Noborio,H., Yoshida,Y., Sohmura,T.
Abstract:
In this paper, we develop a dental surgical system based on mixed reality, which a dental doctor can scrape a concave tooth with complicated shape by a bar located at the tip of turbine. In this system, we represent a tooth and a dental bar as an octree (a set of voxels) and sets of points, respectively. Based on the octree's hierarchical structure in positioning, we quickly detect an intersection between octree-based tooth and point-based bar. Moreover, according to the intersection set, we scrape a tooth by a bar while making force and moment. Finally, many doctors flexibly pick up visual and tactile parameters according to a lot of their experiences. In addition, our system automatically evaluates a student operation against a professional one by comparing their scraping tooth shapes. For this reason, dental students can learn many kinds of surgical operations on demand via PC and internet.
Ukrop,L., Jakubeci,M., Kapec,P.
Abstract:
Data visualization of large abstract data sets and complicated relations is a complex research area with different problems and constraints. Often simple shapes and structures are not very eye-pleasing. Visualization metaphors, which create a mapping between a well-known problem domain and a new complicated problem domain, can produce interesting visualizations. In this paper we propose two metaphorical visualizations of graphs and multidimensional data: we propose a metaphor of soap bubble clusters to visualize graphs and a nebulae (sky) metaphor, which uses nebulae and stars to visualize graphs and multidimensional data.
Horn,M.E.
Abstract:
Special relativity can be modelled mathematically with complex quaternions. The relation between this quaternionic special relativity and spacetime algebra will be discussed from a didactical perspective showing the intrinsic relations between quaternion matrices, Pauli matrices, and Dirac matrices.
Degirmenci,M., Ashyralyev,S.
Abstract:
As outer space image acquisition techniques progress, larger amounts of planetary data sets become available. Impact crater statistics about planets is an important resource as use of this information reveals geological history. Since manual detection of impact craters requires substantial human resource, there is a compelling need to investigate automated crater detection algorithms. In this study, we develop a novel framework to detect Martian impact craters by fusing data obtained from Mars Global Surveyor. In our proposed method, extracted craters from Mars Digital Image Model (MDIM) are crosschecked by using Mars Digital Elevation Model (MDEM). Multi population genetic algorithm (MPGA) has been devised to extract craters from scale invariant feature set found by SIFT algorithm. In order to decrease the number of false positives, extracted from MDIM are validated by detected basins from MDEM. Experimental results on NASA databases suggest high crater detection rates.
Yuanjie,L., Hongbo,L.
Abstract:
The objective of this paper is to discuss an approach which combines B-spline patches and transfinite interpolation to establish linear algebraic system for solving partial differential equations. We modified the Web-spline method developed by Klaus Hollig. First of all, in this new approach, we replace the R-function method with transfinite interpolation to build the function which vanishes on boudaries. Secondly, we simulate the partial differential equation through directly applying differential operators to basis functions, which is similar to the RBF method rather than Hollig's method. Compared to standard FEM, our meshless approach uses regular grids and direct differentiation, thus the time-consuming process for generating meshes and computation of integration is avoided.
Skala,V.
Abstract:
The paper describes a new approach for barycentric coordinates computation using projective space and principle of duality. Duality is also used for formulation of intersection of two planes computation using Plucker coordinates. The approach is convenient especially for CUDA/GPU implementation.