Ciencia y Tecnologías Físicas >
Centro de Física Miguel A. Catalán - Instituto de Óptica (CFMAC-IO) >
(CFMAC-IO) Tesis >
Open Access item New contributions to the multidimensional analysis of signals and images through the pseudo-Wigner Distribution
|Other Titles:||Nuevas contribuciones en el análisis multidimensional de señales e imágenes mediante la pseudo-distribución de Wigner|
|Authors:||Gabarda Tébar, Salvador|
|Advisor:||Cristóbal Pérez, Gabriel|
|Keywords:||Wigner Distribution, Image analysis|
|Abstract:||The subject of this thesis is to develop new applications in signal processing based on
the multidimensional analysis that can be achieved by means of the Wigner distribution
(WD). The kind of signals that have been mainly handled are the signals used in computer vision, as images, or image sequences, without excluding other kind of signals as time series.
Typically, the WD requires a great amount of calculations. To simplify these calculations, a new set of algorithms for 1-D and 2-D signals is introduced in this thesis. The performance of this methodology relays in the use of a directional 1-D discrete version of the WD, along with a small data window, instead of a full 2-D implementation. This technique, added to the continuously increasing speed of the new
processors, approximates the use of the WD to a real-time tool for digital image processing. Comparative measurements are also presented that show how the results obtainable through this analysis, based on a determined and discrete version of the WD, are similar and sometimes better than other classical methods.
Additionally, the multi-directional analysis introduced in this thesis makes possible
the evaluation of the anisotropic nature of the images. This anisotropic evaluation has been used to develop an application for isotrigon textures discrimination and to derive an objective anisotropic measure for image quality assessment. Finally, the methods for image processing developed in this thesis have been
extended successfully to other frames as biomedical signals.|
|Description:||Tesis doctoral de la Facultad de Ciencias (Departamento de Física Matemática y de Fluidos) de la Universidad Nacional de Educación a Distancia (UNED) y del Instituto de Óptica "Daza de Valdés" del Consejo Superior de Investigaciones Científicas (IO-CSIC).-- 166 págs.|
|Appears in Collections:||(CFMAC-IO) Tesis|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.