|
|
Zobecněný váleček
Haindl, Michal ; Hatka, Martin
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 texture 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 purpose of real-time rendering of any type of static or dynamic textures.
|
|
|
Výběr nejinformativnějších proměnných ve statistickém rozpoznávání
Pudil, Pavel ; Somol, Petr ; Haindl, Michal
The research report gives an overview of feature selection techniques in statistical pattern recognition with particular emphasis to methods developed by the researchers participating in MATEO Centre of Mechatronics project. Besides discussing the advances in methodology it attempts to put them into a taxonomical framework. The methods discussed include the latest variants of the optimal algorithms, enhanced sub-optimal techniques and the simultaneous semi-parametric probability density function modelling and feature space selection method. Some related issues are illustrated on real data by means of the Feature Selection Toolbox software.
|
|
| |
|
|
Úvod do statistického rozpoznávání
Pudil, Pavel ; Somol, Petr ; Haindl, Michal
Pattern recognition problem is briefly characterized as a process of machine learning. Its main stages (dimensionality reduction and classifier design) are stated. Statistical approach is given priority here. Two approaches to dimensionality reduction, namely feature selection (FS) and feature extraction (FE) are specified. Though FS is a special case of FE, they are very different from a practical viewpoint and thus must be considered separately.
|
|
|
Rozpoznávání založené na vícerozměrných modelech
Haindl, Michal ; Pudil, Pavel ; Somol, Petr
This chapter explains general model-based approaches to several basic pattern recognition applications followed by a concise description of three fundamental multi-dimensional data model classes. For each model class a solution to parameter estimation and model data synthesis is outlined. Finally an overview of the strengths and weaknesses of studied multi-dimensional data model groups is given.
|
|
| |
|
| |
|
|
Nový algoritmus hledání cesty pro sešívání obrazů a pokročilé dlaždicování textur
Somol, Petr ; Haindl, Michal
We propose a fast and adjustable sub-optimal path search algorithm for finding minimum error boundaries between overlapping images. The algorithm may serve as an alternative to traditional slow path search algorithms like the dynamical programming. We use the algorithm in combination with novel adaptive blending to stitch image regions. The technique is then exploited in a framework for sampling-based texture synthesis where the learning phase is clearly separated and the synthesis phase is very simple.
|
|
|
Váleček - rychlá metoda syntézy textur založená na vzorkování
Haindl, Michal ; Hatka, M.
This paper describes a method for synthesizing natural textures that realistically matches given colour texture appearance. The novel texture synthesis method, which we call the roller, is based on the overlapping tiling and subsequent minimum error boundary cut. An optimal double toroidal patch is seamlessly repeated during the synthesis step. While the method allows only moderate texture compression it is extremely fast and easily implementable in a graphical hardware for purpose of real-time rendering.
|
|
|
Multispectral texture segmentation
Mikeš, Stanislav ; Haindl, Michal
An efficient and robust type of unsupervised multispectral texture segmentation method is presented. The algorithm starts with spectral factorization of an input multispectral texture image using the Karhunen-Loeve expansion. Monospectral factors of single texture patches are assumed to be modelled using a Gaussian Markov random field model. The texture segmentation is done by K-means algorithm in the Markov model parameter space evaluated for each pixel centered image window.
|
guest ::
