Ismail Jouny

The bispectrum of complex signals: definitions and properties

The symmetry properties and the relationship between all forms of third-order cumulants of complex signals are investigated. It is shown that all cumulants (for different position of the complex conjugate) are related by simple transformations. Autoregressive modeling of complex-valued signals using third-order cumulants is also investigated. It is shown that modeling of complex-valued signals requires a different approach from modeling of real-valued signals. >

On SVM for classification of real and synthetic radar signatures

This paper focuses on radar target identification using support vector machines (SVM). The radar features used in this study are impulse response features representing the down range profile of the target as seen by stepped frequency radar. The purpose of this paper is to shed additional light on the benefits of SVM in radar target identification (RTI) under various scenarios of adversity that are commonly addressed in the RTI literature. This paper attempts to maximize the performance of SVM in RTI but does not introduce new SVM kernels, or SVM training methods. The focus is on defining the rewards of SVM in target identification assuming a classifier that is presented with time domain signatures representing the target impulse response at a certain azimuth angle. In particular this paper focuses on assessing the SVM classifier performance in different scenarios, which are discussed in this paper

Clutter suppression and feature extraction for land mine detection using ground penetrating radar

In this paper, we discuss approaches for processing ground penetrating radar (GPR) data for land mine detection. We discuss two methods of clutter suppression using ARMA modeling and also examine feature extraction using both ARMA modeling and complex natural resonance (CNR) modeling. The algorithms presented have been tested on laboratory GPR data and been found to be promising techniques. In future work, the feature vectors produced by these techniques will be processed through appropriate detection algorithms.

Music DOA estimation with compressive sensing and/or compressive arrays

This paper examines the pros and cons of compressed sensing in estimating the direction of arrival (DOA) of sparse signals impinging on an antenna array. The observed signal received by each antenna element may be compressed (temporally [2]) assuming that it is sparse (a valid assumption when dealing with real radar target signature), and the arrays receiving these signals may also be compressed (spatially) by compressing the array manifold from M antennas to Ms antennas according to some random sampling scheme [1,3]. The performance of both approaches is assessed via simulations involving real and synthetic radar scattering signatures obtained using stepped frequency radar. Of particular interest is the impact on array resolution and the ability of compressively sampled antenna array to resolve correlated arrivals.

Target identification in a MIMO environment

A MIMO radar system with M transmitters and N receivers is used for target identification of unknown non-cooperative radar targets. The radar system is assumed wideband with sufficient bandwidth to obtain a target impulse response that shows significant scattering mechanisms and correlates well with the geometry of the target. Contributions in this paper are threefold: First, target identification in a Multi-Input-Multi-Output radar framework is established. Second, the system relies on real scattering signatures as seen by a wideband radar system. Third, the system is tested using real radar signatures collected in a compact range environment using a single-antenna stepped frequency radar interrogating commercial aircraft models at various azimuth angles.

Blind Source Separation in Tumor Detection in Mammograms

Blind source separation (BSS) is a signal processing technique to isolate what information came from a particular source. We apply BSS to mammography in order separate a tumor from the surrounding tissue, making it easier to identify tumors automatically.

Target description using wavelet transform

A method for characterizing radar signatures using the wavelet transform is developed based on the principle of scattering centers. The adopted target representation is based on the colored bright spot approach, which is closely related to wavelet analysis. It is shown how to estimate the number, location, level of integration or differentiation, and scattering intensity of point scatterers using wavelet decomposition. The proposed target analysis method is tested using experimentally measured radar signals.<<ETX>>

Radar backscatter analysis using fractional Fourier transform

This paper focuses on analyzing radar backscatter returns using the fractional Fourier transform. This study is motivated by two factors: first, to examine the radar backscatter mechanism of standard small targets; and second, to extract pertinent scattering features that can be used in target recognition. Radar returns have been examined using time-frequency analysis techniques, particularly those targets with dispersive scattering behavior. The FrFT scattering analysis scheme is tested using real radar signatures of commercial aircraft recorded in the UHF range.

Temporal watermarking invertible only through source separation

This paper presents an application of blind signal separation to the field of digital signatures and watermarking of audio signals. The second-order blind identification (SOBI) algorithm presented by Belouchrani et al. can be used to retrieve a watermark or to identify a signature. [1]. Multiple source signals (one of which acts as a digital signature) are arbitrarily mixed with grossly different proportions and transmitted. Only the linear mixtures are observed and no information is known about the prior mixing. This process allows for digital verification of the original message, as well as the ability to transmit hidden, buried messages.

Neural network face recognition using wavelets

The recognition of human faces is a phenomenon that has been mastered by the human visual system and that has been researched extensively in the domain of computer neural networks and image processing. This research is involved in the study of neural networks and wavelet image processing techniques in the application of human face recognition. The objective of the system is to acquire a digitized still image of a human face, carry out pre-processing on the image as required, an then, given a prior database of images of possible individuals, be able to recognize the individual in the image. The pre-processing segment of the system includes several procedures, namely image compression, denoising, and feature extraction. The image processing is carried out using Daubechies wavelets. Once the images have been passed through the wavelet-based image processor they can be efficiently analyzed by means of a neural network. A back- propagation neural network is used for the recognition segment of the system. The main constraints of the system is with regard to the characteristics of the images being processed. The system should be able to carry out effective recognition of the human faces irrespective of the individuals facial-expression, presence of extraneous objects such as head-gear or spectacles, and face/head orientation. A potential application of this face recognition system would be as a secondary verification method in an automated teller machine.