Search academic texts

Information × Registration Number 0223U004783, 0118U001789 , R & D reports Title Numerical methods for solving problems in mechanics and computer vision popup.stage_title Head Verbitskyi Viktor V., Кандидат фізико-математичних наук Registration Date 04-12-2023 Organization Odessa I. I. Mechnikov National University popup.description2  The stress-strain state of the half-strip under the influence of steady oscillations was investigated, the characteristics of stresses and displacements were determined depending on the frequency of loading and the size of the section. A scheme of the finite element method for constructing a numerical solution of the boundary value problem for the Helmholtz equation with mixed Dirichlet and Robin boundary conditions is proposed. The finite element approximation of the problem is built on the basis of the variational formulation of the boundary value problem. The basis functions of the finite-dimensional subspaces of the space of the variational problem are constructed using the solutions of the Dirichlet boundary value problem for the homogeneous equation on the finite element. This result can be interpreted as a manifestation of the superconvergence of the finite element method. On the other hand, the scheme avoids the "contamination" effect, when the exact and finite element solutions differ significantly at the grid nodes. The conducted computational experiment confirms the accuracy of the scheme. It has been established that it is best to use methods based on autoregression to solve the problem of detecting and removing impulse noise in audio signals. A method has been described and developed that can be used to normalize audio signals that have been corrupted by impulse noise such as clicks, pops, or scratches. When an audio signal is input, the system automatically finds distorted samples and replaces them with more appropriate values. This algorithm is a simple but effective method of restoring sound signals damaged by impulse noise. The limited range of parameters and the fact that the detection and interpolation phases have common parameters allow fast signal processing in automatic mode. Results obtained on a variety of signals (classical, jazz, voice, etc.) demonstrate that a fully automated system with a well-defined set of parameters can achieve good p Product Description popup.authors Verbitskyi Viktor V. Zhuravlyova Zinaida Y. Moroz Volodymyr V. Reut Viktor V. Tairova Maria S. popup.nrat_date 2023-12-04 Close