4th International Workshop on New Approaches for

Multidimensional Signal Processing

NAMSP 2023

Technical University of Sofia, Sofia, Bulgaria

July 06-08, 2023

---

Plenary Speakers

Prof. Szilvia Nagy, Széchenyi István University, Győr, Hungary

Title of Lecture:

Structural entropies in image processing

Abstract: Entropies help to describe, characterize and classify probability distributions. In the case of image processing and other 2-dimensional signal processing, there are multiple approaches to use Shannon/von Neumann and Rényi entropies. Pipek and Varga introduced their structural entropy on electron densities. Their defini-tion is based on Rényi entropies, and it is used to characterize the localization type of the electron density in 1-, 2-, and 3-dimensional systems. However, other mul-tidimensional signals can be normalized in a way, that they could be interpreted as probability distribution, thus this toolbox can be used on images as well, both as a filter, and as a characterization method. There are other approaches to use entro-py for characterizing structure of graphs, granular systems, or distributions with subsystems of various size or other structural properties. In this paper the similar-ities and differences of these two approaches are studied together with their appli-cation possibilities.

Biographical Notes: Szilvia Nagy received her PhD at the Budapest University of Technology and Eco-nomics in Physics in 2005. She is a full professor at the Széchenyi István University since 2019, at the Department of Telecommunications. Her main research interest includes electron structures, signal and image analysis, wavelets and entropies.