Search

Pipelines play a crucial role in today's infrastructure. However, they are easily affected by environmental conditions. Therefore, regular inspections are essential to ensure the integrity of pipelines. This study introduces a new approach that combines variation mode decomposition (VMD) and empirical wavelet transform (EWT) with continuous wavelet transform (CWT) for acoustic data to improve the signal-to-noise ratio (SNR) and accurately detect pipeline leakage location in the time domain. The acoustic data was gathered in three distinct scenarios inside an open environment, where external noise considerably influenced the original signal.VMD and EWT are employed to enhance the signal-to-noise ratio (SNR). VMD efficiently distinguishes high-frequency noise and assures distinct mode separation, whilst EWT delivers precise temporal localization for non-stationary signals. Based on the results of VMD and EWT, certain IMFs were selected for further study and SNR is computed among the respective chosen IMFs and original signals to measure the improvement. Subsequently, the CWT was utilized on the chosen IMF for feature extraction, effectively finding leakage locations in the time domain. This study illustrates the feasibility of successfully mitigating noise and extracting signals in open and noisy environments, thus enhancing the accuracy of leak detection. The results establish a robust basis for subsequent investigations in pipeline leak detection and demonstrate the applicability of advanced signal processing approach to real-time leak detection systems. [ABSTRACT FROM AUTHOR]

Interference mitigation has been a vital area of focus within the realm of automotive radar research, holding paramount importance for augmenting advanced driver-assistance capabilities, particularly in vehicle-dense scenarios. This paper presents a novel interference mitigation scheme based on continuous wavelet transform and morphological component analysis (CWT-MCA). This method, in essence, transforms the time-domain signal into the time-frequency domain, subsequently enabling the direct separation of target and interference components, thereby circumventing the necessity for interference detection and identification. Additionally, this paper introduces a series of MCA algorithms grounded in various transform domains, which are subsequently compared with the traditional linear prediction based approach. We show through comprehensive numerical simulations and real-data experiments that CWT-MCA algorithm exhibits markedly superior interference mitigation performance while offering lower computational complexity, which makes it well-suited for real-time implementation in automotive radar systems. • Continuous wavelet transform and morphological component analysis are applied to mitigate interference for automotive radars. • Fast algorithm is proposed to accelerate computation. • Numerical studies and real-data experiments are conducted to validate the algorithm performance. [ABSTRACT FROM AUTHOR]

• Detailed analysis of constant-Q filterbank based time-frequency representations for speech emotion recognition. • Constant-Q representations show higher time invariance at low frequencies and vice versa. • Better pitch resolution than mel-scaled representations is given by Constant-Q representations. • Constant-Q based non-linearity outperforms mel-scale with various neural network classifiers. This work analyzes the constant-Q filterbank-based time-frequency representations for speech emotion recognition (SER). Constant-Q filterbank provides non-linear spectro-temporal representation with higher frequency resolution at low frequencies. Our investigation reveals how the increased low-frequency resolution benefits SER. The time-domain comparative analysis between short-term mel-frequency spectral coefficients (MFSCs) and constant-Q filterbank-based features, namely constant-Q transform (CQT) and continuous wavelet transform (CWT), reveals that constant-Q representations provide higher time-invariance at low-frequencies. This provides increased robustness against emotion irrelevant temporal variations in pitch, especially for low-arousal emotions. The corresponding frequency-domain analysis over different emotion classes shows better resolution of pitch harmonics in constant-Q-based time-frequency representations than MFSC. These advantages of constant-Q representations are further consolidated by SER performance in the extensive evaluation of features over four publicly available databases with six advanced deep neural network architectures as the back-end classifiers. Our inferences in this study hint toward the suitability and potentiality of constant-Q features for SER. [ABSTRACT FROM AUTHOR]

Abstract: This paper investigates two new multisensor data fusion algorithms for object detection in monitoring of industrial processes. The goals were to reduce the rate of false detection and obtain reliable decisions on the presence of target objects. The monitoring system uses acceleration sensors and is used as a sensor-cluster. In principle the approach can include arbitrary data acquisition techniques. Two approaches were proposed. The first uses a short-time Fourier transform (STFT) as a prefilter to extract relevant features from the acceleration signals. The features extracted from different sensor channels are first classified using support vector machine (SVM)-based filters. A novel decision fusion process to combine individual decisions was developed. The second approach uses a continuous wavelet transform (CWT) as a prefilter to extract relevant features from the acceleration signals. The features extracted from different sensor signals are subjected to further prefiltering processes before SVM-based classification. The individual decision functions are then combined in a decision fusion module. The classification system was trained and validated using real industrial data. The two approaches were tested using the same data and their performance and modeling complexity are compared. The developed approaches show strong improvements in detection and false alarm rates. [Copyright &y& Elsevier]

Abstract: This paper presents a novel approach for the local analysis of the contour of a planar real world shape. The 1D representation of that contour is a very complicated signal with several non isolated and oscillating singularities, which represent the micro-structure of the shape. The analysis of such a signal is usually difficult because of the presence of spurious phenomena in its time-scale representation, typical of oscillating singularities. The aim of the proposed model is to exploit the time-scale behavior of the energy of wavelet coefficients to extract a particular sequence of scales where those spurious phenomena are strongly reduced. The locality of the wavelet transform is then used to segment the contour into meaningful subregions, in agreement with their local spectral properties. Experimental results show that the proposed model overcomes some limits of existing methods for the analysis of real-world shapes micro-structure. [Copyright &y& Elsevier]