Pradip Sircar

Fast communication: EEG signal analysis using FB expansion and second-order linear TVAR process

In this paper, we propose a second-order linear time-varying autoregressive (TVAR) process for parametric representation of the electroencephalogram (EEG) signals. The coefficients of the Fourier-Bessel (FB) series expansion have been used to constitute a feature vector for segmentation of the EEG signal. Our approach is novel in the sense that by selecting an appropriate data length, we find a simple model for parametric representation of the EEG signals. The complete method for estimation of model parameters is presented in this work.

Analysis of multicomponent AM-FM signals using FB-DESA method

The discrete energy separation algorithm (DESA) together with the Gabors filtering provides a standard approach to estimate the amplitude envelope (AE) and the instantaneous frequency (IF) functions of a multicomponent amplitude and frequency modulated (AM-FM) signal. The filtering operation introduces amplitude and phase modulations in the separated monocomponent signals, which may lead to an error in the final estimation of the modulation functions. In this paper, we have proposed a method called the Fourier-Bessel expansion-based discrete energy separation algorithm (FB-DESA) for component separation and estimation of the AE and IF functions of a multicomponent AM-FM signal. The FB-DESA method does not introduce any amplitude or phase modulation in the separated monocomponent signal leading to accurate estimations of the AE and IF functions. Simulation results with synthetic and natural signals are included to illustrate the effectiveness of the proposed method.

A novel approach for automated detection of focal EEG signals using empirical wavelet transform

The determination of epileptogenic area is a prime task in presurgical evaluation. The seizure activity can be prevented by operating the affected areas by clinical surgery. In this paper, an automatic approach has been presented to detect electroencephalogram (EEG) signals of non-focal and focal groups. The proposed approach can be used to determine the area linked to the focal epilepsy. In our method, the EEG signal is decomposed into rhythms using empirical wavelet transform technique. The two-dimensional (2D) projections of the reconstructed phase space (RPS) have been obtained for the rhythms. Area measures for various RPS plots are estimated using central tendency measure (CTM) parameter. The area parameters are used with least-squares support vector machine (LS-SVM) classifier to classify the focal and non-focal classes of EEG signals. In this work, we have achieved a maximum classification accuracy of 90%, sensitivity and specificity of 88 and 92%, respectively, using 50 pairs of focal and non-focal EEG signals. The same method has achieved maximum classification accuracy, sensitivity and specificity of 82.53, 81.60 and 83.46%, respectively, with 750 pairs of signals. The developed prototype can be used for the epileptic patients and aid the clinicians to confirm diagnosis.

Complex AM signal model for non-stationary signals

Abstract A novel signal model consisting of a weighted sum of complex amplitude modulated signals is proposed as a suitable representation for non-stationary signals like speech. The estimation of model parameters is carried out by utilizing the accumulated autocorrelation functions of the modelled signal. The developed model is first fitted on noise-corrupted synthesized data, and then on sampled voiced speech data. The study demonstrates the suitability of the model.

Content Based Color Image Classification using SVM

We propose a novel approach for content based color image classification using Support Vector Machine (SVM). Traditional classification approaches deal poorly on content based image classification tasks being one of the reasons of high dimensionality of the feature space. In this paper, color image classification is done on features extracted from histograms of color components. The benefit of using color image histograms are better efficiency, and insensitivity to small changes in camera view-point i.e. translation and rotation. As a case study for validation purpose, experimental trials were done on a database of about 500 images divided into four different classes has been reported and compared on histogram features for RGB, CMYK, Lab, YUV, YCBCR, HSV, HVC and YIQ color spaces. Results based on the proposed approach are found encouraging in terms of color image classification accuracy.

Speech Analysis using Fourier-Bessel Expansion and Discrete Energy Separation Algorithm

In this paper, a new technique based on the Fourier-Bessel (FB) expansion is presented for separating multiple formants of a speech signal. The discrete energy separation algorithm (DESA) is applied to an isolated speech formant to extract the instantaneous frequency (IF) and the time-varying amplitude envelope (AE) of the formant. It is demonstrated that the proposed technique which is called the FB-DESA technique is a powerful tool for speech formant analysis. The technique is based on simple principle and it is easy to implement

A novel technique to reduce cross terms in the squared magnitude of the wavelet transform and the short-time Fourier transform

The energy distribution of a non-stationary signal is often represented by the squared magnitude of the wavelet transform (WT) or the short-time Fourier transform (STFT). The energy distribution contains cross terms, which can cause problems while analyzing the multicomponent signals. In this paper, a unified approach is developed for the time-frequency analysis of a multi-component signal without introducing cross terms. The Fourier-Bessel (FB) series expansion of the multicomponent signal is used for separating and reconstructing the individual components of the signal, and then, the time-frequency energy distributions are computed for the individual components. The simulation results show that the proposed approach based on the FB series expansion is a powerful tool for analyzing multicomponent signals

Modified Exponential Companding for PAPR Reduction of OFDM Signals

In this paper, a new nonlinear companding technique, called the modified exponential companding technique, is proposed to reduce the peak to average power ratio (PAPR) of the orthogonal frequency division multiplexed (OFDM) signals. Instead of transforming the Rayleigh distributed OFDM signal into uniformly distributed signal as in the case of exponential companding proposed by Jiang, et al. (2005), we consider different probability distributions of the transformed signal, by introducing a control parameter a, which reduces the PAPR further. Simulation study demonstrates that the proposed method results in better PAPR reduction and improved BER performance compared to the exponential companding.

System identification from nonuniformly spaced signal measurements

Abstract A couple of techniques for parameter estimation of a signal consisting of complex exponentials are presented. The system model employs the higher-order derivatives, or zero-initial conditioned integrals of the signal, together with the signal values. When the signal is sampled at nonuniformly distributed points, the orthogonal polynomial approximation and minimum error-variance criterion are used to compute all the values needed in the system models. The developed system models are demonstrated to give results better in accuracy than what can be obtained by employing Pronys method in a specific problem.

Time-frequency analysis using time-order representation and Wigner distribution

A new method for time-frequency analysis of a signal, which combines the time-order representation and the Wigner-Ville distribution (WVD) has been presented in this paper. The time-order representation based on short-term Fourier-Bessel (FB) expansion, decomposes a multicomponent signal into a number of monocomponent signals, and then the WVD technique is applied on each component of the composite signal to analyze its time-frequency distribution (TFD). Simulation results with real bat signal are included to illustrate the effectiveness of the proposed method.