Discrete Wavelet Transform Decomposition Research Articles

Recognition of Multiple Power Quality Disturbances Based on Discrete Wavelet Transform and Improved Long Short-Term Memory Networks

In the past few years, distributed electricity generation from renewable sources, or microgrid systems, has been connected to the grid to increase power supply stability. This responds to government policy regarding commitment to using 100% renewable energy in operations (RE100) project efforts. This results in the entry of power electronic or non-linear equipment into the electrical system, making it more sensitive. Moreover, multiple power quality disturbances (PQDs) consist of a variety of single disturbances. Analysis of complex multi-label patterns is a challenging task. In this paper, we propose a methodology to address this challenge by leveraging Discrete Wavelet Transform (DWT) and Improved Long Short-Term Memory Networks (LSTM). Firstly, multiple PQDs are synthesized utilizing a mathematical model based on IEEE standards 1159-2019. Secondly, the obtained PQDs are decomposed into nine feature classes, yielding detailed (cDs) and approximation (cAs) coefficients through Five-Level DWT Decomposition. Furthermore, we conducted a comparative analysis of each component across five different wavelet functions: haar1, db4, bior1.3, coif2, and sym4. Thirdly, the cDs and cAs coefficients derived from each wavelet type undergo statistical analysis before being inputted into the LSTM model for classification of each feature class. Our results highlight that cD5 components obtained from the db4 wavelet exhibit the highest accuracy rate of 93.86%. This finding elucidates the significance of selecting appropriate wavelet types and compositions for the successful classification of multiple PQDs.

EMG Instrumentation Modeling and Feature Processing Based On Discrete Wavelet Transform

Electromyography (EMG) instrumentation is essential in generating electrical signals from skeletal muscles. EMG sensors are helpful in various cases requiring the detection of human muscle contractions, neuromuscular disorders, and rehabilitation. EMG instrumentation is divided into two parts, namely, the analogue part and the digital part. The EMG instrumentation design comprises a digital-to-analog converter (DAC), instrumentation amplifier, filter, and analog-to-digital converter (ADC). Meanwhile, in digital signal processing adopting the Discrete Wavelet Transform (DWT) method, frequency analysis using DWT has proven superior. It is used in various research and has exceptionally detailed coefficient features for classifying neuromuscular disease signals. Therefore, this research aims to design analogue and digital EMG instrumentation and identify features in the form of detailed coefficients. The data used are two Physionet signals from the anterior tibialis body with myopathy and neuropathy disorders. The results obtained for EMG analogue instrumentation provide the expected results until they reach the filter component stage. The resulting signal forms a block in the filter component, different from the initial EMG signal. Meanwhile, the DWT decomposition results are of the Daubechies4 wavelet type with the highest level 17, which has a high detail coefficient at low frequencies, high dilation and the result of a mixture of neuropathy and myopathy EMG signals, or in other words, at low energies, this result is by the DWT theorem. Determining the efficiency of the DWT detailed coefficient feature requires further study with the same signal subject. The DWT features obtained can then be developed for various needs in EMG signal recognition.

Wavelet decomposition-based multi-stage feature engineering and optimized ensemble classifier for stock market prediction

Expanded feature engineering to handle complex and noisy stock data is not much explored in financial forecasting. Along with feature engineering, hyperparameter tuning is also important. This paper provides a solution to handle these issues by elaborating on Discrete Wavelet Transform (DWT)-based feature engineering and hyperparameter-tuned ensemble model. The Multi-Stage Feature Engineering (MSFE) is proposed in which DWT-based decomposition is utilized to handle the noise. DWT decomposition expands the features; therefore, two-stage feature reduction is proposed in which first the filter method is used, and then the probabilistic method is utilized. Next, hyperparameter tuning of the ensemble model is offered through Particle Swarm Optimization (PSO). The proposed model is named as Wavelet-Particle Swarm Optimization (WPSO). WPSO is tested and evaluated on the three stock indices (NIFTY, NASDAQ, and NYSE), and provided 92.51%, 94.18%, and 87.62% accuracy for NIFTY, NASDAQ, and NYSE, respectively. The WPSO is validated by comparing it with state-of-the-art methods. The performance of the WPSO is statistically analyzed through the Bonferroni–Dunn post hoc test where WPSO positioned at rank 1 for all the evaluation metric and datasets. The WPSO empirically verifies that improving feature quality through MSFE and hyperparameter tuning of ensemble model significantly improves the predictive outcomes.

Expert diagnostic system for detection of hypertension and diabetes mellitus using discrete wavelet decomposition of photoplethysmogram signal and machine learning technique

Hypertension is a leading risk factor for cardiovascular disease (CVD) and the overlap of diabetes mellitus (DM) with hypertension can lead to severe complications. There is a need for early diagnosis and risk stratification to implement an overall risk management strategy. Presently, the conventional method is not suitable for large-scale screening. The primary aim of this study is to develop an automated diagnostic system that uses Photoplethysmogram (PPG) signals for the non-invasive diagnosis of hypertension and DM-II. The proposed model uses a statistical feature extracted by decomposing the PPG signal up to level 11 into a sub-band using Discrete wavelet transform (DWT), and a variety of classifiers are used for the classification of hypertension and detection of DM-II patients. Three feature selection techniques used are Spearman correlation, ReliefF and Minimum Redundancy Maximum Relevance (mRMR) to select 20 top features out of 130 features using correlation with systole blood pressure (SBP), diastole blood pressure (DBP) values and D-II. The highest accuracy attained by the Adaptive neural fuzzy system (ANFIS) for classification categories such as normal (NT) vs prehypertension(PHT), NT vs. hypertension type 1 (HT-I), NT vs hypertension type 2 (HT-II) in terms of F1 score are 92.%, 98.5%, 98.3% (SBP) and 83.1%, 95.6%, 86.8% (DBP),respectively. The accuracy achieved by the adaptive-network-based fuzzy inference system (ANFIS) for the classification of normal (non-diabetic) vs. diabetic patients is 99.1%. The hybrid learning algorithm-based classifier achieved higher accuracy for hypertension risk stratification as compared to the hard computing classifier, which requires parameter tuning and DWT decomposition is robust to the noisy signal, overcoming the limitation of the morphological feature-based model.

Grading diabetic retinopathy using multiresolution based CNN

Diabetic Retinopathy (DR) refers to a medical condition that affects the eye; it occurs due to diabetes, and, if not detected early on, results in a reduction of visual capacity and may even result in blindness. The process of diagnosing DR manually by ophthalmologists can be both time-consuming and expensive, and there is a risk of misdiagnosis. On the other hand, computer-aided diagnostic systems can provide a more accurate and efficient diagnosis and can help ophthalmic specialists by offering a second opinion for effective treatments. This paper presents a method in which the feature extraction is based on the multiresolution-based decomposition of Discrete Wavelet Transform (DWT), and classification is performed using Convolutional Neural Network (CNN) for grading DR images. The suggested approach begins by utilizing Contrast Limited Adaptive Histogram Equalization (CLAHE) to refine the contrast level of the fundus images as a pre-processing technique. The images in the datasets (IDRiD, DDR, and EyePACS) are not balanced, so oversampling is applied to ensure that an equal number of images from each of the grade categories are present during the training process. The results of the experiments demonstrate that the proposed model achieves a classification accuracy of 90.07%, 96.20%, and 93.53% for all DR stages, outperforming existing models on all the three datasets. Therefore, the proposed method offers a superior alternative to current approaches.

Adaptive Wavelet De-noising Algorithm using Absolute Difference Optimization Technique for Partial Discharge Signal

Discrete Wavelet Transform (DWT) de-noising method is widely used for one-dimension partial discharge (PD) signals measured from medium voltage underground cable. However, DWT de-noising has several drawbacks that prevent the DWT de-noising from improving its de-noising effectiveness In DWT de-noising, the two most important parameters are decomposition level and mother wavelet. The aforementioned parameters must be varied according to the noise level in the measured PD signal in order to effectively suppress the noise of the measured PD signal. In this paper, an adaptive DWT de-noising algorithm based on the Absolute Difference Optimizing (ADO) technique is presented to effectively suppress the varying noise levels in measured PD signal. First, the measured PD signal will be de-noised using a Daubechies 3 (db3) mother wavelet and a DWT decomposition level ranging from 1 to 10. Second, the de-noised PD signal will be subjected to the ADO technique. The sum of the absolute difference of local maxima in the de-noised PD signal will be used as an indicator to select the best decomposition level for the de-noised PD signal. Finally, the best-selected de-noised PD signal by using the ADO technique will be used to estimate the PD location on the underground cable. The results of PD location error using the ADO technique and normal DWT de-noising will be compared. The findings show that the ADO-based adaptive DWT de-noising algorithm significantly improved the de-noising process of the measured PD signal.

Speckle Noise Reduction in Ultrasound Images

Abstract: Ultrasound Imaging is used to examine various organs in the Human Body. However, in the process of obtaining a Ultrasound Image, Speckle noise is generated due to backscattered echo signals. So, a noise reduction method is required. To improve the speckle noise reduction, we propose a novel algorithm using characteristics of speckle noise and filtering methods based on Speckle Reducing Anisotropic Diffusion (SRAD) filtering, Discrete Wavelet Transform (DWT), Weighted Guided Image Filtering (WGIF) and Gradient Domain Guided Image Filtering (GDGIF). The SRAD filter is exploited as a preprocessing filter because it can be directly applied to a medical US image containing speckle noise without a log-compression. The wavelet domain has the advantage of suppressing the additive noise. Therefore, a homomorphic transformation is utilized to convert the multiplicative noise into additive noise. After two-level DWT decomposition is applied, to suppress the residual noise of an SRAD filtered image, GDGIF and WGIF are exploited to reduce noise from seven high-frequency subband images and one low-frequency sub-band image, respectively. Finally, a noise-free image is attained through inverse DWT and an exponential transform

Remove Blur Image Using Bi-Directional Akamatsu Transform and Discrete Wavelet Transform

Purpose: Image is an imitation of everything that can be materialized, and digital images are taken using a machine. Although digital image capture uses machines, digital images are not free from interference. Image restoration is needed to restore the quality of the damaged image.Methods: Bi-directional Akamatsu Transform is proven to have an effective performance in reducing blur in images. Meanwhile, Discrete Wavelet Transform has been widely used in digital image processing research. We had been investigated the image restoration method by combining Bi-directional Akamatsu Transform and Discrete Wavelet Transform. Bi-directional Akamatsu Transform applied in Low-Low (LL) sub-band is the Discrete Wavelet Transform decomposition image most similar to the original image before decomposing. In this study, there are still shortcomings, including the determination of the values of N, up_enh, and down_enh, which are still manual. Manually setting the three values makes the Bi-directional Akamatsu Transform method not get the best results. With the use of machine learning methods can get better restoration results. Further testing is also needed for a more diverse and robust blur. The image data has a resolution of 256x256, 512x512, and 1024x1024. The image will be directly converted to a grey-scale image. The converted image will be given an attack model: average blur, gaussian blur, and motion blur. The image that has been attacked will apply two restoration methods: the proposed method and the Bi-direction Akatamatsu Transform. These two restoration images will then be compared using PSNR.Result: The average PSNR value from the restoration of the proposed method is 0.1446 higher than the average PSNR value from the restoration of the Bi-directional Akamatsu Transform method. When we compare it with the average PSNR value of the Akamatsu Transform restoration method, the average PSNR of the proposed method is 0.2084.Value: The combination of DWT and akamatsu transform results produce good PSNR values even though they have gone through the blurring method in image restoration.

A Watermarking Optimization Method Based on Matrix Decomposition and DWT for Multi-Size Images

Image watermarking is a key technology for copyright protection, and how to better balance the invisibility and robustness of algorithms is a challenge. To tackle this challenge, a watermarking optimization method based on matrix decomposition and discrete wavelet transform (DWT) for multi-size images is proposed. The DWT, Hessenberg matrix decomposition (HMD), singular value decomposition (SVD), particle swarm optimization (PSO), Arnold transform and logistic mapping are combined for the first time to achieve an image watermarking optimization algorithm. The multi-level decomposition of DWT is used to be adapted to multi-size host images, the Arnold transform, logistic mapping, HMD and SVD are used to enhance the security and robustness, and the PSO optimized scaling factor to balance invisibility and robustness. The simulation results of the proposed method show that the PSNRs are higher than 44.9 dB without attacks and the NCs are higher than 0.98 under various attacks. Compared with the existing works, the proposed method shows high robustness against various attacks, such as noise, filtering and JPEG compression and in particular, the NC values are at least 0.44% higher than that in noise attacks.

Fault diagnosis for vehicle air conditioning blower using deep learning neural network

This study presents a fault diagnosis system for vehicle heating, ventilation and air conditioning (HVAC) acoustic signal with various feature extractions in deep learning neural network. Traditionally, sound used for fault diagnosis or signal classification is observed the difference of energy in time or frequency domains. Unfortunately, the frequency smearing effect often arises in some critical conditions. In the present study, discrete wavelet transform (DWT) and wavelet packet transform (WPT) are proposed in fault diagnosis. Meanwhile, when using mechanical learning methods, the data are relatively large, in order to reduce the amount of data, DWT and WPT low-frequency decomposition could be used to improve the performance. Furthermore, the signal characteristics more comprehensive, this study attempts to use the feature extraction method of wavelet packet conversion to improve the signal characteristics. In the experiment process, the operation state of the blade blower in the vehicle air conditioner, four different faults were designed, test database was established through sound to classify, and identify the data using deep neural networks to achieve the purpose of blower fault diagnosis. In data analysis, the original signal is presented through wavelet packet decomposition and discrete packet conversion technology, compared with traditional time and frequency domain signals to explore the identification rate, identification speed and related issues. Experimental results show that using WPT combined with deep neural networks have good fault diagnosis and discrimination capabilities, training, and identification time is shorter than time-frequency domain signals.

Wavelet transform approach for denoising and decomposition of satellite-derived ocean color time-series: Selection of optimal mother wavelet

Wavelet Transform (WT) has remarkable advantages for feature extraction and denoising of satellite-derived ocean color time series datasets. The selection of mother wavelet (MWT) is the main challenging issue in the wavelet transform analysis of ocean color time series, since different MWT applied to the same dataset may produce different results. In this study, 57 MWTs were investigated to find the optimal wavelet functions for WT analysis of ocean color time series datasets with particular emphasis on the quantitative approaches. The daily, weekly, and monthly averaged of MODIS Level-3 Chlorophyll-a (Chl-a) series were denoised and decomposed by Discrete Wavelet Transform (DWT) using different MWTs. The seasonal trends of all datasets were extracted using the Seasonal Trend decomposition procedure based on LOcally wEighted Scatterplot Smoothing (LOESS) (STL) and Census X-11 methods, used as references. The statistical Cross Correlation Function (CCF), Signal to Noise Ratio (SNR), and Mean Standard Error (MSE), indexes, plus a Magnitude-Shape quantitative model were used to calculate the correlation and similarity between original and corresponding decomposed and denoised reference Chl-a signals. The similarity indexes were analyzed to evaluate the similarity and correlation between reconstructed denoised signals with original Chl-a series, and between DWT seasonal components with the reference seasonal trends. For a more assessment, two different weekly averaged Chl-a series of SeaWiFS satellite sensor over Bermuda-Atlantic region (open ocean waters) and Chesapeake Bay (coastal waters) were evaluated using SeaBASS in situ measurements, and seasonal trends in comparison with corresponding DWT components. The results revealed that the “db10” and “db5” from Daubechies wavelet family were respectively the best options for denoising and decomposition of Chl-a series using WT methods. Also MWTs from Biorthogonal and Reverse biorthogonal wavelet families gave a poor performance, and therefore these functions are not recommended to use in DWT denoising and decomposition of Chl-a signals.

Wavelet spectral analysis and attribute ranking applied to automatic classification of power quality disturbances

The increasing use of non-linear loads and devices in power systems has increased the incidence of waveform changes in electrical energy signals. The intensity and frequency of these disturbances define energy quality being delivered to the consumers. This work explored such issue and proposed two new approaches to improve the effectiveness of automatic classification of these disorders. The first concerns the extraction of attributes. Initially, the best correlation between the Discrete Wavelet Transform (DWT) decomposition levels and the original signals was investigated, pointing out the 8th level as the most representative. Regarding the disturbances and their respective 8th level detail coefficients, 7 traditional metrics were applied and a new feature was proposed: the weighted center frequency (wcf). Additionally, a spectral analysis of the 8th level detail coefficients was performed and thus 4 new features completed a vector with 12 attributes. The second novelty consisted of a multiple indexes analysis to define a global marker for ranking and selecting attributes. To the best of our knowledge, this is the first time that a multi-parametric ranking technique has been used to support automatic classification of electrical disturbances. These features were extracted from a synthetic database of 10,000 signals and subjected to different classifiers: an Artificial Neural Network (ANN) and a Support Vector Machine (SVM). The results showed that two proposed attributes related to the spectral analysis of the DWT decomposition had a relevant impact on the performance of the classifiers. Furthermore, the results of the classifiers confirmed the correctness of the ranking method by finding a difference of about 30% when comparing the classification performances considering half of the highest ranked attributes against the lowest ranked ones. Finally, the results obtained were compared with those of other prominent works in the literature, such that its relevance was verified by pointing to effective solutions for the automatic classification of electrical disturbances.

LWKPCA: A New Robust Method for Face Recognition Under Adverse Conditions

Over the last two decades, face recognition (FR) has become one of the most prevailing biometric applications for effective people identification as it offers practical advantages over other biometric modalities. However, current state-of-the-art findings suggest that FR under adverse and challenging conditions still needs improvements. This is because face images can contain many variations like face expression, pose, and illumination. To overcome the effect of these challenges, it is necessary to use representative face features using feature extraction methods. In this paper, we present a new feature extraction method for robust FR called Local Binary Pattern and Wavelet Kernel PCA (LWKPCA). The proposed method aims to extract the discriminant and robust information to minimize recognition errors. This is obtained first by the best use of nonlinear projection algorithm called RKPCA. Then, we adapted the algorithm to reduce the dimensionality of features extracted using the proposed Color Local Binary Pattern and Wavelets transformation called Color LBP and Wavelet Descriptor. The general idea of our descriptor is to find the best representation of face image in a discriminant vector structure by a novel feature grouping strategy generated by the Three-Level decomposition of Discrete Wavelet Transform (2D-DWT) and Local Binary Pattern (LBP). Extensive experiments on four well-known face datasets namely ORL, GT, LFW, and YouTube Celebrities show that the proposed method has a recognition accuracy of 100% for ORL, 96.84% for GT, 99.34% for LFW, and 95.63% for YouTube Celebrities.

Carotid Sound Signal Artifact Detection based on Discrete Wavelet Transform Decomposition

Abstract Auscultation methods allow a non-invasive diagnosis of cardiovascular diseases like atherosclerosis based on blood flow sounds of the carotid arteries. Since this process is highly dependent on the clinician’s experience, it is of great interest to develop automated data processing techniques for objective assessment. We have recently proposed a computerassisted auscultation system that we use to acquire carotid blood flow sounds. In this work, we present an approach for detecting artifacts within the blood flow sound caused by swallowing or coughing events. For this purpose, we first decompose the signal using a discrete wavelet transform (DTW). Then, we compute an energy ratio between the DWT scales associated with the signal information with and without artifacts using a sliding window of 1 s length. Evaluation based on Kruskal-Wallis and Wilcoxon rank-sum tests shows a statistically significant difference (p-value<.0001) between the signal with and without artifact. Therefore, the proposed method allows the identification of the studied signal artifacts.

Visualization of blood vessels in in vitro raw speckle images using an energy-based on DWT coefficients

The visualization and localization of blood vessels is an important task to determine the presence and the health status of microvasculature in the biological tissue. Laser Speckle Contrast Imaging is one of the most widely employed techniques to study blood vessels; even so, it has some drawbacks in analyzing deep blood vessels (>100μm) since the image noise level increases.The Wavelet Approach is a model of automatic denoising for contrasted in vitro Raw Speckle images using an energy criterion. The criterion selects the more suitable denoising level from the Discrete Wavelet Transform decomposition using the detail coefficients. Then, the segmentation of low-noise images by mathematical morphology techniques establish the blood vessel and biological tissue location. Finally, the region corresponding to the blood vessel and the low-noise images are used to improve the visualization of blood vessels.Results show that a Wavelet Approach improves the visualization of blood vessels up to a depth of 400μm. Furthermore, the proposed model demonstrates that the automatic denoising criterion improves the localization of superficial (≤100μm) and deep (>100μm) blood vessels.

Wavelet Transform and Fractal Theory for Detection and Classification of Self-extinguishing and Fugitive Power Quality Disturbances

In this paper, fractal theory and wavelet transform are combined to detect and classify self-extinguishing and fugitive scenarios of power quality disturbances (PQDs). After deciding whether the disturbance is simple or complex, the additional voltage is denoised through Discrete Wavelet Transform (DWT); the denoising process is adapted according to whether the distorted voltage contains oscillatory transients or not. At the detection stage, the grille fractal dimension of the DWT decomposition detail is computed. Then, a threshold is deduced to detect the start and end moments of the disturbance. The results reveal that the proposed detection scheme yields accurate location of PQDs even in the presence of high oscillatory transients. An algorithm based on geometric and statistical approaches is developed at the classification stage to recognize PQDs automatically. The geometric classification is based on Continuous Wavelet Transform (CWT), whereas the statistical classification is based on Multifractal Detrended Fluctuations Analysis (MFDFA) and an energy metric. The results prove that the combination of geometric and statistical classification can serve as an effective discrimination tool for PQDs. The major strength of the proposed approach is its ability to interpret the impact of each disturbance on the multifractal behavior of the nominal voltage, thus giving the possibility to draw the necessary generalizations for real-time applications.

Visually Meaningful Image Encryption Scheme Based on DWT and Schur Decomposition

A visually meaningful image encryption scheme with an asymmetric structure based on the discrete wavelet transform (DWT) and Schur decomposition is proposed in this study. First, the RSA algorithm is used to generate the initial values for the chaotic system to produce the random sequence. Then, both scrambling and diffusion operations are performed on the plain image to obtain the preencrypted image. Moreover, the Schur decomposition is applied on the preencrypted plain image to obtain the upper triangular and orthogonal matrices. Second, the cover image is scrambled followed by a DWT operation. Four subbands are then formed, namely, LL, HL, LH, and HH. Finally, the former upper triangular matrix and orthogonal matrix are embedded into subbands LH and HH produced by the cover image. After the application of the inverse DWT and inverse scrambling operation, the final visually meaningful cover image embedded with a secret plain image can be obtained. No one can identify any useful information about the plain image from the final embedded cover image, nor can anybody know that there is any hidden secret image. Experimental simulations show that the normalized correlation values between the original cover image and the final visually meaningful cover image are approximately 0.9997. Therefore, the proposed encryption scheme is imperceptible for secret image communications.

Detection of Induction Motor Improper Bearing Lubrication by Discrete Wavelet Transforms (DWT) Decomposition

The present paper deals with healthy and improper bearing lubrication signals analysis using Discrete Wavelet Transform (DWT) enhanced by MATLAB/ Wavelets toolbox analysis. The identification of bearing faults from the time or the frequency domain are difficult due to non stationary vibration signal. Therefore, for more accurate faults information and identification of bearing with lubrication defects (improper or absence of lubrication), the DWT is used. The validation of this procedure is conducted by an experimental setup designed for vibration signal acquisition and the complete analysis is finalized by MATLAB/ Wavelets toolbox. The recorded data used for the validation are the signals of healthy and un-lubricated bearing driven at a rotation speed of 1500 rpm by 0.78 KW three phase induction motor. From the obtained results it can be observed that, for medium speeds DWT decomposition enhanced by MATLAB Wavelets Toolbox procedure is efficient for improper lubricated bearing related faults diagnosis and detection.

Wavelet feature selection of audio and imagined/vocalized EEG signals for ANN based multimodal ASR system

Human-Machine Interaction (HMI) systems demand the use of multiple modalities for correct interaction. Research on these systems started with audio signals for speech recognition and now progressing towards co-operation of other biosignals. Thus, the paper presents an Automatic Speech Recognition (ASR) system based on a single and multiple modalities that include audio and Electroencephalogram (EEG) signals to explore speech recognition. It extracts speech information concealed in audio and ten channels of imagined EEG (EEG-i) & vocalized EEG (EEG-v) signals. Three Wavelet Transform (WT) methods - Discrete Wavelet Transform (DWT), Wavelet Packet Decomposition (WPD) & hybrid of DWT & WPD (DWPD) with four-level decomposition is used to transform the signals into WT coefficients. Then, six statistical parameters are computed from WT coefficients to generate 63 (26-1) feature vectors for each method. An exhaustive search from 63 feature vectors is conducted to determine the best parameter combination that attains good accuracy with ANN classifier. Then, accuracy is improved with 5 level decomposition on WPD coefficients along with the best parameter combination. Results include the accuracy of unimodal ASR & multimodal ASR. WPD method achieved best accuracy as 74.48%, 56.29%, 42.02%, 77.97% & 78.90% for multiclass classification of prompts+words based on audio, EEG-i, EEG-v, audio + EEG-i & audio + EEG-v respectively. It indicates that speech recognition is possible from EEG signals & the fusion of audio with EEG enhances the recognition rate of audio & EEG. The results also show that the proposed method outperforms other methods in the area.

Despeckling Algorithm for Removing Speckle Noise from Ultrasound Images

Ultrasound (US) imaging can examine human bodies of various ages; however, in the process of obtaining a US image, speckle noise is generated. The speckle noise inhibits physicians from accurately examining lesions; thus, a speckle noise removal method is essential technology. To enhance speckle noise elimination, we propose a novel algorithm using the characteristics of speckle noise and filtering methods based on speckle reducing anisotropic diffusion (SRAD) filtering, discrete wavelet transform (DWT) using symmetry characteristics, weighted guided image filtering (WGIF), and gradient domain guided image filtering (GDGIF). The SRAD filter is exploited as a preprocessing filter because it can be directly applied to a medical US image containing speckle noise without a log-compression. The wavelet domain has the advantage of suppressing the additive noise. Therefore, a homomorphic transformation is utilized to convert the multiplicative noise into additive noise. After two-level DWT decomposition is applied, to suppress the residual noise of an SRAD filtered image, GDGIF and WGIF are exploited to reduce noise from seven high-frequency sub-band images and one low-frequency sub-band image, respectively. Finally, a noise-free image is attained through inverse DWT and an exponential transform. The proposed algorithm exhibits excellent speckle noise elimination and edge conservation as compared with conventional denoising methods.