An algorithm of the sequence of artificial symmetric signals for comparing and creating a new convolution method

Abstract

The objective of this article is to create an algorithm of the sequence of artificial signals that can be used to compare and create methods for processing one- and two-dimensional signals. It will then be implemented to compare feature extraction methods that rely on discrete wavelet transforms. The discrete wavelet transform is superior to other signal processing techniques in several ways. Developing a feature set is a crucial step in using the discrete wavelet transform. Mean value and standard deviation are suggested as feature extraction techniques in this study. The mean value is the only option selected for the first feature extraction method; the mean value and standard deviation are selected for the second feature extraction method. To build any number of artificial signal sequences from a single, several conditions are taken into account, for example, their symmetry, they are supposed to be located at the same distance from each other, that is, with an equal step. Symmetrical signal sequences constructed in this way differ from common wellknown signal sequences, such as Fourier series, in that they converge to a given signal in equal steps.