Wavelet-based Filtering Research Articles

Wavelet-based Filter Bank Approach for Image Compression and Enhancement in Alzheimer\'s Disease Diagnosis Using Medical Imaging

Abstract: This research presents a wavelet-based filter bank approach for image compression and enhancement in the context of Alzheimer's disease diagnosis using medical imaging. The proposed system leverages the PyWavelets and OpenCV libraries in Python to implement Discrete Wavelet Transform (DWT) and Inverse Discrete Wavelet Transform (IDWT) algorithms. The primary objective is to develop an efficient image compression technique while preserving visual quality and enhancing relevant features for improved diagnostic accuracy. The system utilizes filter banks to process wavelet coefficients, enabling selective enhancement of patterns or biomarkers associated with Alzheimer's disease. Results show that the decompressed images closely resemble the originals, with low Mean Squared Error (MSE) and high Peak Signal-to-Noise Ratio (PSNR) values. Visual inspection and quantitative metrics confirm the successful preservation of image quality and structural details. The wavelet coefficients visualization highlights the captured significant features crucial for reconstruction

Enhancing road traffic flow prediction with improved deep learning using wavelet transforms

Precise traffic flow prediction is a central component of advancing intelligent transportation systems, providing essential insights for optimizing traffic management, reducing travel times, and alleviating congestion. This study introduces an efficient deep learning approach that synergistically integrates the benefits of wavelet-based denoising and Recurrent Neural Networks (RNNs). This integrated methodology is introduced to effectively capture the inherent nonlinearity and temporal dependencies in time series traffic data. Specifically, Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) are introduced to address the challenges associated with accurately forecasting traffic flow. To enhance prediction quality, traffic flow data is preprocessed using exponential smoothing and wavelet-based filtering as denoising filters, effectively eliminating outliers. The effectiveness of the proposed techniques is evaluated using traffic measurements collected from diverse highway locations across California, including the Old Bayshore highway, situated south of Interstate 880 (I880), and the Ashby Ave highway, positioned west of Interstate 80 (I80) in the San Francisco Bay Area. The results obtained through integrating both architectures, including LSTM and GRU, within the wavelet transform-based filter demonstrate an enhancement in forecasting performance. Specifically, LSTM with wavelet-based denoising using Symlet and Haar wavelets achieved high prediction performance with an average R2 of 0.982 and 0.9811, respectively.

Performance analysis of remote photoplethysmography deep filtering using long short-term memory neural network

BackgroundRemote photoplethysmography (rPPG) is a technique developed to estimate heart rate using standard video cameras and ambient light. Due to the multiple sources of noise that deteriorate the quality of the signal, conventional filters such as the bandpass and wavelet-based filters are commonly used. However, after using conventional filters, some alterations remain, but interestingly an experienced eye can easily identify them.ResultsWe studied a long short-term memory (LSTM) network in the rPPG filtering task to identify these alterations using many-to-one and many-to-many approaches. We used three public databases in intra-dataset and cross-dataset scenarios, along with different protocols to analyze the performance of the method. We demonstrate how the network can be easily trained with a set of 90 signals totaling around 45 min. On the other hand, we show the stability of the LSTM performance with six state-of-the-art rPPG methods.ConclusionsThis study demonstrates the superiority of the LSTM-based filter experimentally compared with conventional filters in an intra-dataset scenario. For example, we obtain on the VIPL database an MAE of 3.9 bpm, whereas conventional filtering improves performance on the same dataset from 10.3 bpm to 7.7 bpm. The cross-dataset approach presents a dependence in the network related to the average signal-to-noise ratio on the rPPG signals, where the closest signal-to-noise ratio values in the training and testing set the better. Moreover, it was demonstrated that a relatively small amount of data are sufficient to successfully train the network and outperform the results obtained by classical filters. More precisely, we have shown that about 45 min of rPPG signal could be sufficient to train an effective LSTM deep-filter.

A data-driven Gaussian process filter for electrocardiogram denoising

Gaussian Processes (GP)-based filters, which have been effectively used for various applications including electrocardiogram (ECG) filtering can be computationally demanding and the choice of their hyperparameters is typically ad hoc.We develop a data-driven GP filter to address both issues, using the notion of the ECG phase domain -- a time-warped representation of the ECG beats onto a fixed number of samples and aligned R-peaks, which is assumed to follow a Gaussian distribution. Under this assumption, the computation of the sample mean and covariance matrix is simplified, enabling an efficient implementation of the GP filter in a data-driven manner, with no ad hoc hyperparameters. The proposed filter is evaluated and compared with a state-of-the-art wavelet-based filter, on the PhysioNet QT Database. The performance is evaluated by measuring the signal-to-noise ratio (SNR) improvement of the filter at SNR levels ranging from -5 to 30dB, in 5dB steps, using additive noise. For a clinical evaluation, the error between the estimated QT-intervals of the original and filtered signals is measured and compared with the benchmark filter.It is shown that the proposed GP filter outperforms the benchmark filter for all the tested noise levels. It also outperforms the state-of-the-art filter in terms of QT-interval estimation error bias and variance.The proposed GP filter is a versatile technique for preprocessing the ECG in clinical and research applications, is applicable to ECG of arbitrary lengths and sampling frequencies, and provides confidence intervals for its performance.

Evaluation of Highway Hydroplaning Risk Based on 3D Laser Scanning and Water-Film Thickness Estimation.

Hydroplaning risk evaluation plays a pivotal role in highway safety management. It is also an important component in the intelligent transportation system (ITS) ensuring human driving safety. Water-film is the widely accepted vital factor resulting in hydroplaning and thus continuously gained researchers’ attention in recent years. This paper provides a new framework to evaluate the hydroplaning potential based on emerging 3D laser scanning technology and water-film thickness estimation. The 3D information of the road surface was captured using a vehicle-mounted Light Detection and Ranging (LiDAR) system and then processed by a wavelet-based filter to remove the redundant information (surrounding environment: trees, buildings, and vehicles). Then, the water film thickness on the given road surface was estimated based on a proposed numerical algorithm developed by the two-dimensional depth-averaged Shallow Water Equations (2DDA-SWE). The effect of the road surface geometry was also investigated based on several field test data in Shanghai, China, in January 2021. The results indicated that the water-film is more likely to appear on the rutting tracks and the pavement with local unevenness. Based on the estimated water-film, the hydroplaning speeds were then estimated to represent the hydroplaning risk of asphalt pavement in rainy weather. The proposed method provides new insights into the water-film estimation, which can help drivers make effective decisions to maintain safe driving.

Filter group delays equalization for 2D discrete wavelet transform applications

This paper presents a new approach to perform near-perfect image reconstruction for an online/real-time two-dimensional discrete wavelet transform (2D-DWT) analysis/synthesis applications, such as image enhancement, compression, and watermarking. The proposed new model is based on equalizing the cumulative filter group delays over all calculation paths by inserting appropriate delays. This model has been implemented in MATLAB/SIMULINK and compared to the default model using six wavelet-based filter sets and two dyadic decomposition/reconstructed levels. Peak signal-to-noise ratios (PSNRs) of images reconstructed by the proposed equalized model present an improvement of more than 60 dB, compared to those of images reconstructed by the default model.

A semi-analytical solution and AI-based reconstruction algorithms for magnetic particle tracking.

Magnetic particle tracking is a recently developed technology that can measure the translation and rotation of a particle in an opaque environment like a turbidity flow and fluidized-bed flow. The trajectory reconstruction usually relies on numerical optimization or filtering, which involve artificial parameters or thresholds. Existing analytical reconstruction algorithms have certain limitations and usually depend on the gradient of the magnetic field, which is not easy to measure accurately in many applications. This paper discusses a new semi-analytical solution and the related reconstruction algorithm. The new method can be used for an arbitrary sensor arrangement. To reduce the measurement uncertainty in practical applications, deep neural network (DNN)-based models are developed to denoise the reconstructed trajectory. Compared to traditional approaches such as wavelet-based filtering, the DNN-based denoisers are more accurate in the position reconstruction. However, they often over-smooth the velocity signal, and a hybrid method that combines the wavelet and DNN model provides a more accurate velocity reconstruction. All the DNN-based and wavelet methods perform well in the orientation reconstruction.

Improved Process Monitoring Scheme Using Multi-Scale Independent Component Analysis

The task of fault detection is crucial in modern chemical industries for improved product quality and process safety. In this regard, data-driven fault detection (FD) strategy based on independent component analysis (ICA) has gained attention since it improves monitoring by capturing non-gaussian features in the process data. However, presence of measurement noise in the process data degrades performance of the FD strategy since the noise masks important information. To enhance the monitoring under noisy environment, wavelet-based multi-scale filtering is integrated with the ICA model to yield a novel multi-scale Independent component analysis (MSICA) FD strategy. One of the challenges in multi-scale ICA modeling is to choose the optimum decomposition depth. A novel scheme based on ICA model parameter estimation at each depth is proposed in this paper to achieve this. The effectiveness of the proposed MSICA-based FD strategy is illustrated through three case studies, namely: dynamic multi-variate process, quadruple tank process and distillation column process. In each case study, the performance of the MSICA FD strategy is assessed for different noise levels by comparing it with the conventional FD strategies. The results indicate that the proposed MSICA FD strategy can enhance performance for higher levels of noise in the data since multi-scale wavelet-based filtering is able to de-noise and capture efficient information from noisy process data.

Multiresolution classification of turbulence features in image data through machine learning

During large-scale simulations, intermediate data products such as image databases have become popular due to their low relative storage cost and fast in-situ analysis. Serving as a form of data reduction, these image databases have become more acceptable to perform data analysis on. We present an image-space detection and classification system for extracting vortices at multiple scales through wavelet-based filtering. A custom image-space descriptor is used to encode a large variety of vortex-types and a machine learning system is trained for fast classification of vortex regions. By combining a radial-based histogram descriptor, a bag of visual words feature descriptor, and a support vector machine, our results show that we are able to detect and classify vortex features at various sizes at multiple scales. Once trained, our framework enables the fast extraction of vortices on new, unknown image datasets for flow analysis.

Ambient suppression in vibration bump test using wavelet-based filter

Modal analysis is generally used to determine the structural dynamic characteristics of a system through experimental. The main purpose of a modal analysis is to identify the modal parameter, which consists of natural or resonant frequencies, mode shapes and damping. However, this technique required all the system in shut down condition in order to perform modal testing. The presence of ambient force such as motor unbalanced can cause errors in measurement of time response. Therefore, a study on how to improve the quality of the acquired signals in EMA is crucial in order to increase the efficiency of EMA technique under the presence of ambient force. This paper introduces a method to eliminate the ambient component in measured vibration response using the wavelet-based filter. The two experiments have been carried involving two different conditions, shutdown condition and operating condition. The sources of the ambient in operating condition are induced from a motor and speed controller cooling fan. The proposed method utilizes the discrete wavelet transform (DWT) and improvised spectral subtraction in filtering the ambient. The discrete wavelet transform (DWT) is applied to both ambient data and total response data to decompose the signal by a factor of two into several levels of the wavelet decomposition. The modification of the discrete wavelet transform (DWT) method has been done where the wavelet thresholding is replaced with the spectral subtraction to suppress the ambient in the signal. The spectral subtraction is used to suppress the ambient in each wavelet coefficient at each level of decomposition to avoid the losses of useful signal data while filtering the ambient. The decomposed wavelet signal is reconstructed and the result is being compared with the baseline data. The Frequency Response Function obtained from the reconstructed signal shows the harmonics feature from ambient excitation were successfully suppressed. The results show that the proposed approach is effective to suppress the ambient effect in vibration response measurement.

Leakage Detection In Pipeline Using Wavelength

Nowadays natural gas transport and distribution is a complex and currently growing and increasing. Natural gas produced from well need to transport in a great distance before reaching it point of use. The pipeline is designed to quickly and efficiently transport the gas from its origin to the high demand area. Either pipelines transportation for water supply or natural gas, leakage is unacceptable problem. Small leak along the pipeline is hard to detect. The objective of this study is to build the test rig galvanized iron and MDPE pipelines. Besides that, the main objective is to determine the leak detection in gas pipeline using wavelet-based filtering. . Main point of each journal is compared in order to determine the problems arise from the previous research. It is then follow by the methodology which will discuss further in this chapter. From methodology, it is known that the data taken can be analysing through Daisy Lab and Math lab software Wavelet and cross correlation is used to analyse the signal in Matlab. From the result, it show that the leak can be identified based on the peak of amplitude of the signal. The result for galvanized iron pipe is not acceptable due to short pipeline length. Thus it can be concluded that leak can be determined using wavelet-based filtering. As the conclusions, the propose technique can be used to determine the leak in pipelines

Experimental investigation of impact forces induced by plunging breakers on a vertical cylinder

A plunging breaker propagates with different stages, each of which induces a different force on structures. In this study, physical experiments of the different stages of a plunging breaker impacting a vertical cylinder are implemented in a wave flume. The concept of decomposing the measured forces into quasi-static forces and dynamic forces is adopted. The method proposed by Irschik et al. (2004) using an FFT filter and EMD (Empirical Mode Decomposition) is applied in the decomposition process. Then a new method with a continuous wavelet-based filter is used that produces a more reasonable result. Subsequently, the results of the measured forces, the quasi-static forces, and the dynamic forces are analyzed by considering the different stages of the plunging breaker. The results show the peak forces induced by the various stages of the plunging breaker are quite different. Prior to the overturning jet impinging the free surface, the quasi-static forces due to the various stages of the plunging breaker are approximately the same, while the dynamic forces contribute to different strengths of the force oscillations and exhibit a similar response. Following the mixture of water with air, the peak value of the quasi-static force is no longer invariant but declines sharply.

ПРИМЕНЕНИЕ ТЕОРИИ ВЕЙВЛЕТОВ ПРИ СЖАТИИ И ФИЛЬТРАЦИИ ГЕОИНФОРМАЦИИ

The purpose of the article is to develop a detailed and accessible technology for the application of wavelets in the processing of geo-information, the subject of research is wavelet-based filtering and compression of geo-information. The research methodology is based on the modern theory of wavelets in the light of linear algebra. Research methods involve study and generalization, abstraction, formalization, mathematical modeling using computer programs compiled by the authors.After the introduction and formulation of the problem, the basic positions of linear algebra are presented, on which the content of the article is based when constructing orthonormal bases in one- and two-dimensional cases. First, the application of the general theory to the decomposition of the vector of initial data in the Haar and Shannon bases is given. Further, on the basis of the Haar basis, orthonormal bases of wavelet transforms and filtering information are constructed. The procedure for creating wavelet filters by a sequence of convolutions, the use of MSA analysis for constructing an orthonormal basis of the wavelet transform is considered. Implemented the practical possibility of wavelet filtering based on specific programs for modeling geo-information data fields and images, data compression and filtering. The result of the work is the methods of constructing orthonormal bases by various methods of wavelet transform, based on which algorithms and corresponding computer programs for geoinformation compression are compiled using the example of terrain and photographic images. The efficiency of geoinformation compression and noise filtering using wavelets was investigated. A method has been developed for determining the value of a filter depending on the accuracy of the initial geo-information, illustrated by the example of calculating the filter value for compressing information about the heights of the terrain. The same technique is recommended for image filtering.

TFAW: Wavelet-based signal reconstruction to reduce photometric noise in time-domain surveys

Context. There have been many efforts to correct systematic effects in astronomical light curves to improve the detection and characterization of planetary transits and astrophysical variability. Algorithms such as the trend filtering algorithm (TFA) use simultaneously-observed stars to measure and remove systematic effects, and binning is used to reduce high-frequency random noise. Aims. We present TFAW, a wavelet-based modified version of TFA. First, TFAW aims to increase the periodic signal detection and second, to return a detrended and denoised signal without modifying its intrinsic characteristics. Methods. We modified TFA’s frequency analysis step adding a stationary wavelet transform filter to perform an initial noise and outlier removal and increase the detection of variable signals. A wavelet-based filter was added to TFA’s signal reconstruction to perform an adaptive characterization of the noise- and trend-free signal and the underlying noise contribution at each iteration while preserving astrophysical signals. We carried out tests over simulated sinusoidal and transit-like signals to assess the effectiveness of the method and applied TFAW to real light curves from TFRM. We also studied TFAW’s application to simulated multiperiodic signals. Results. TFAW improves the signal detection rate by increasing the signal detection efficiency (SDE) up to a factor ∼2.5× for low S/R light curves. For simulated transits, the transit detection rate improves by a factor ∼2 − 5× in the low-S/R regime compared to TFA. TFAW signal approximation performs up to a factor ∼2× better than bin averaging for planetary transits. The standard deviations of simulated and real TFAW light curves are ∼40% better compared to TFA. TFAW yields better MCMC posterior distributions and returns lower uncertainties, less biased transit parameters and narrower (by approximately ten times) credibility intervals for simulated transits. TFAW is also able to improve the characterization of multiperiodic signals. We present a newly-discovered variable star from TFRM.

Evaluation of filters over different stimulation models in evoked potentials

Filtering is a key process which removes unwanted parts of signals. During signal recording, various forms of noises distort data. Physiological signals are highly noise sensitive and to evaluate them powerful filtering approaches must be applied. The aim of this study is to compare modern filtering approaches on scalp signals. Brain activities were generally examined by brain signals like EEG and evoked potentials (EP). In this study, data were recorded from university students whose age between 18 and 25 years with visual and auditory stimuli. Discrete wavelet transforms, singular spectrum analysis, empirical mode decomposition and discrete Fourier transform based filters were used and compared with raw data on classification performance. Higuchi fractal dimension and entropy features were extracted from EEG; P300 features were extracted from EP signals. Classification was applied with support vector machines. All filtered data gave better scores than raw data. Empirical mode decomposition (EMD) and Fourier-based filter yielded lower results than the discrete wavelet-based filter. Singular spectrum analysis gave the best result at 84.32%. The current study suggests that singular spectrum analysis removes noise from sensitive physiological signals, and EMD requires new mode selection procedures before resynthesizing.

Regional and residual gravity anomaly separation using the singular spectrum analysis-based low pass filtering: a case study from Nagpur, Maharashtra, India

We present here a singular spectrum analysis (SSA)-based low pass filtering algorithm for regional and residual gravity anomaly separation. The data adaptive decomposition in SSA frequency filtering algorithm facilitates reduction of the artefacts in the filtering of the non-linear and non-stationary gravity data. Initially, the method was tested on synthetic gravity data representing combined response of regional and residual components derived from the geological structures like infinite horizontal layers, intrusive dykes, deep-seated faults and volcanic intrusive bodies and compared with fast Fourier transform (FFT) wavenumber and wavelet-based filtering techniques. The results show that SSA-based filtered output exhibits a better match with pure synthetic data than the output generated from the FFT and wavelet filtering methods. The underlying method was then applied to two parallel gravity profiles of real data from the Umred coalfield, Nagpur, Maharashtra, India. Further, the SSA-based filtered regional anomaly was modelled using Geosoft GM-SYS software to construct the model of crustal structure of the study region. In essence, the modelling results suggest the following: (i) a basin-type graben structure with variable trap and sedimentary thicknesses and (ii) deep seated faults on either sides of the basin. These results correlate fairly well with known regional geology attesting the authenticity of the regional models generated from the two gravity profiles, which also agree well with each other. We, therefore, conclude that the SSA-based filtering technique is robust for regional gravity anomaly separation and could be effectively exploited for filtering other geophysical data.

An efficient wavelet-based adaptive filtering algorithm for automatic blind speech enhancement

In this paper, we address the problem of speech enhancement by adaptive filtering algorithms. A particular attention has been paid to the backward blind source separation (BBSS) algorithm and its use in crosstalk resistant speech enhancement applications. In this paper, we propose to implement the BBSS algorithm in the wavelet-domain. The proposed backward wavelet BBSS (WBBSS) algorithm is then used in speech enhancement application when important crosstalk interferences are presents. The new WBBSS algorithm shows better performances in terms of convergence speed and steady state in comparison with the classical BBSS one. The performances properties of the proposed algorithm are evaluated in term of segmental SNR (SegSNR), segmental mean square error (SegMSE), and cepstral distance (CD) criteria. The obtained results have confirmed the best performance of the proposed WBBSS algorithm in a lot of situations when blind noisy observations are available.

Comparing different smoothing methods to detect double-cropping rice phenology based on LAI products – a case study in the Hunan province of China

ABSTRACTMany studies have demonstrated the remarkable potential of assimilating remotely sensing leaf area index (LAI) products into crop models in estimating regional crop yield. To ensure the temporal consistency between crop models and remote-sensing system, it is prerequisite to derive the crop phenology information from the LAI products. However, previous studies mainly detected the phenology through the vegetation index (VI). Although some pieces of research applied LAI in phenology monitoring for trees and shrubs, fewer focused on crops, especially those with two or three growing seasons annually. Thus, which smoothing algorithm methods are suitable to obtain phenology of double-cropping rice and their difference in smoothing for crops are still unknown. Based on the Global Land Surface Satellite (GLASS)LAI products, we applied four favourite smoothing algorithms (Asymmetric Gaussian fitting, Double Logistic fitting, Savitzky–Golay filter, and Wavelet-based Filter method) to reduce noise and reconstruct the LAI profile and then detected the phenological information of double-cropping rice in Hunan Province. Compared with ground actual observations, we found that two fitting methods are not suitable to smooth double-cropping rice LAI, while the wavelet method performed the best. Based on the wavelet method, we estimated the phenological information of double-cropping rice at different regional scales as well and the results reflected that the accuracy of regional estimation is also acceptable. This study implied that the wavelet method is rather suitable to detect phenological information of crops from LAI products, which provides narrow gaps between two growing season. Our contribution can benefit researchers who focus on agriculture or remote sensing, especially those who would like to assimilate remotely sensed information into crop growth models.

Exploring an optimal wavelet-based filter for cryo-ET imaging

Cryo-electron tomography (cryo-ET) is one of the most advanced technologies for the in situ visualization of molecular machines by producing three-dimensional (3D) biological structures. However, cryo-ET imaging has two serious disadvantages—low dose and low image contrast—which result in high-resolution information being obscured by noise and image quality being degraded, and this causes errors in biological interpretation. The purpose of this research is to explore an optimal wavelet denoising technique to reduce noise in cryo-ET images. We perform tests using simulation data and design a filter using the optimum selected wavelet parameters (three-level decomposition, level-1 zeroed out, subband-dependent threshold, a soft-thresholding and spline-based discrete dyadic wavelet transform (DDWT)), which we call a modified wavelet shrinkage filter; this filter is suitable for noisy cryo-ET data. When testing using real cryo-ET experiment data, higher quality images and more accurate measures of a biological structure can be obtained with the modified wavelet shrinkage filter processing compared with conventional processing. Because the proposed method provides an inherent advantage when dealing with cryo-ET images, it can therefore extend the current state-of-the-art technology in assisting all aspects of cryo-ET studies: visualization, reconstruction, structural analysis, and interpretation.

Combined non-parametric and parametric approach for identification of time-variant systems

Abstract Identification of systems, structures and machines with variable physical parameters is a challenging task especially when time-varying vibration modes are involved. The paper proposes a new combined, two-step – i.e. non-parametric and parametric - modelling approach in order to determine time-varying vibration modes based on input-output measurements. Single-degree-of-freedom (SDOF) vibration modes from multi-degree-of-freedom (MDOF) non-parametric system representation are extracted in the first step with the use of time-frequency wavelet-based filters. The second step involves time-varying parametric representation of extracted modes with the use of recursive linear autoregressive-moving-average with exogenous inputs (ARMAX) models. The combined approach is demonstrated using system identification analysis based on the experimental mass-varying MDOF frame-like structure subjected to random excitation. The results show that the proposed combined method correctly captures the dynamics of the analysed structure, using minimum a priori information on the model.