DISCRETE TRANSFORM IN BLOCK FORM AND INVARIANCE TO PROCESSING DIRECTION

Рассмотрен ряд вопросов инвариантности дискретных блочных преобразований. Показано, что смена направления обработки при выполнении обратимых преобразований приводит к изменению, получаемых частотных составляющих. Предложены математические модели блочных дискретных преобразований на основе матричного представления. Предлагается рассматривать дискретные преобразования в зависимости от количества блоков и направления обработки. Приводятся результаты моделирования на основе дискретного преобразования Фурье. Some questions of invariance of a discrete block transforms are considered. It is shown a processing direction change in a reversible transforms performing leads to a frequency components change. Mathematical models are proposed for block discrete transforms based on a matrix representation. The discrete transforms are proposed depending on a blocks number and direction of processing. Modeling results on the discrete Fourier transform are presented.

Comparison of machine learning methods for crack localization

In this paper, the Haar wavelet discrete transform, the artificial neural networks (ANNs), and the random forests (RFs) are applied to predict the location and severity of a crack in an Euler–Bernoulli cantilever subjected to the transverse free vibration. An extensive investigation into two data collection sets and machine learning methods showed that the depth of a crack is more difficult to predict than its location. The data set of eight natural frequency parameters produces more accurate predictions on the crack depth; meanwhile, the data set of eight Haar wavelet coefficients produces more precise predictions on the crack location. Furthermore, the analysis of the results showed that the ensemble of 50 ANN trained by Bayesian regularization and Levenberg–Marquardt algorithms slightly outperforms RF.

An Effective Method to Denoiseeeg, ECG and PPG Signals based on Meyer Wavelet Transform

In this analysis, the prevailing role of the wavelet transform in the interrogation of the ECG is discussed in detail, where both the constant and the discrete transform are considered in turn.A Wavelet denoising is functional on the original signal to eradicate high frequency noise, and then a process based on Meyer wavelet transform combined with adaptive filter is functional to eradicate the motion artifact. This approach uses Meyer Wavelet decomposition to extract the motion artifact, which is subsequently utilized as the reference input of an adaptive filter for noise cancellation. The technique diminishes the overhead of the circuit because it does not need a separate collection of reference input signal which link to noise. Testing results illustrate that this approach can efficiently remove motion artifact and make better the signal quality.

Discrete Transform Based Image Fusion

With the advancement of image processing, the distinct area of image fusion has been explored. The word fusion represents a way of obtaining data acquired in several domains. A technique of merging useful data from input images is defined as image fusion. It improves features and performance. Fused image includes all the important features of input images without introducing any artifacts. This paper depicts the basic of image fusion and fusion techniques. Paper mainly focuses on frequency domain techniques. Image fusion widely used in surveillance, medical diagnosis, biometric, enhanced vision system and remote sensing.