Nidal Kamel

Glove-Based Approach to Online Signature Verification

Utilizing the multiple degrees of freedom offered by the data glove for each finger and the hand, a novel online signature verification system using the singular value decomposition (SVD) numerical tool for signature classification and verification is presented. The proposed technique is based on the Singular value decomposition in finding r singular vectors sensing the maximal energy of glove data matrix A, called principal subspace, so the effective dimensionality of A can be reduced. Having modeled the data glove signature through its r-principal subspace, signature authentication is performed by finding the angles between the different subspaces. A demonstration of the data glove is presented as an effective high- bandwidth data entry device for signature verification. This SVD-based signature verification technique is tested and its performance is shown to be able to produce equal error rate (EER) of less than 2.37 percent.

A survey of methods used for source localization using EEG signals

Abstract The EEG source localization which is used to localize the electrical activity of brain has been an active area of research as it provides useful information for study of brains physiological, mental and functional abnormalities. This problem is called EEG inverse problem. The localization of the active sources needs the solution of ill posed EEG inverse problem. Since the foundation of this field till today, many methods have been developed with the aim of in-depth localization, high resolution, reduction in localization/energy error and decreased computational time. In this survey, EEG inverse problem is discussed with its primary to most developed and recent solutions. The introduction to the field along with the categorization of different solutions is provided. Also, the relative advantages and limitations for each method are discussed. Finally, the challenges and future recommendations are provided, in the end, for further improvement of EEG inverse problem in terms of resolution, computational power and localization error.

Dynamic Signature Verification Using Sensor Based Data Glove

Handwritten signature verification is a well-established and potential area of research with numerous applications such as commercial (e.g., credit card, bank check verification etc.), government (e.g., National ID card, Drivers license, passport control etc.) and forensic (e.g., corpse identification) application. In this paper, we propose a new approach to deal with the problem of handwritten signature verification and forgery detection using data glove. The technique is based on linearly projecting the glove signature into a low-dimensional space, through the singular value decomposition (SVD). The Euclidean distance between the different groups of singular values is used to measure the authenticity of the tried signatures. The reliability and efficiency of the proposed system against forgeries are tested and reported. A comparative analysis has also been shown for data gloves with 14, 5, and 4 sensors respectively

Subspace-Based Technique for Speckle Noise Reduction in SAR Images

Image-subspace-based approach for speckle noise removal from synthetic aperture radar (SAR) images is proposed. The underlying principle is to apply homomorphic framework in order to convert multiplicative speckle noise into additive and then to decompose the vector space of the noisy image into signal and noise subspaces. Enhancement is performed by nulling the noise subspace and estimating the clean image from the remaining signal subspace. Linear estimator minimizing image distortion while maintaining the residual noise energy below some given threshold is used to estimate the clean image. Experiments are carried out using synthetically generated data set with controlled statistics and real SAR image of Selangor area in Malaysia. The performance of the proposed technique is compared with Lee and homomorphic wavelet in terms of noise variance reduction and preservation of radiometric edges. The results indicate moderate noise reduction by the proposed filter in comparison to Lee but with a significantly less blurry effect and a comparable performance in terms of noise reduction to wavelet but with less artifacts. The results also show better preservation of edges, texture, and point targets by the proposed filter than both Lee and wavelet and less required computational time.

Single-Trial Subspace-Based Approach for VEP Extraction

A signal subspace approach for extracting visual evoked potentials (VEPs) from the background electroencephalogram (EEG) colored noise without the need for a prewhitening stage is proposed. Linear estimation of the clean signal is performed by minimizing signal distortion while maintaining the residual noise energy below some given threshold. The generalized eigendecomposition of the covariance matrices of a VEP signal and brain background EEG noise is used to transform them jointly to diagonal matrices. The generalized subspace is then decomposed into signal subspace and noise subspace. Enhancement is performed by nulling the components in the noise subspace and retaining the components in the signal subspace. The performance of the proposed algorithm is tested with simulated and real data, and compared with the recently proposed signal subspace techniques. With the simulated data, the algorithms are used to estimate the latencies of P100, P 200, and P300 of VEP signals corrupted by additive colored noise at different values of SNR. With the real data, the VEP signals are collected at Selayang Hospital, Kuala Lumpur, Malaysia, and the capability of the proposed algorithm in detecting the latency of P100 is obtained and compared with other subspace techniques. The ensemble averaging technique is used as a baseline for this comparison. The results indicated significant improvement by the proposed technique in terms of better accuracy and less failure rate.

Front-end estimation of Noise Power and SNR in OFDM systems

In this paper a front-end noise power and SNR estimation method based on one OFDM preamble is proposed and compared with previously published SNR estimators - none of which are front-end estimator. The estimator makes use of two identical halves property of time synchronization preamble used in OFDM systems and relies on autocorrelation of the same. The current estimator is compared with two existing SNR estimators, the Reddy estimator and Subspace based estimator in terms of normalized mean square error (NMSE) and estimated SNR. It is observed that current estimator gives SNR estimates better than Reddy and subspace estimators. The estimator performs SNR estimation at front-end of the receiver unlike all other estimators which perform SNR estimation at back-end of the receiver.

Routing strategies in hierarchical cluster based mobile wireless sensor networks

Ubiquitous communication networks is a keystone for New Generation Network (NWGN). Mobile Wireless Communication Networks (MWSN) is a viable solution to accomplish the requirements of NWGN. Due to mobility of sensor nodes, the data reliability and end-to-end delay with energy efficiency in the network is an enormous concern. Various real-time and delay sensitive applications enforced to use both environments mobile and fixed sensor nodes, whereas the others claims an entire mobile sensors environments in network. Packet loss ratio and end-to-end delay happened because of the nodes mobility which is directly impact to degrade the quality of service, network lifetime and energy consumption. This paper enlightens a comprehensive comparison between single and multi hop inter-cluster routing strategy from cluster head to base station. Moreover, the performance of multi hop routing is calculated and compared with single hop LEACH routing strategy. The simulation results reveal that multi hop routing strategy is to increase the sensor nodes throughput and network lifetime but not efficient approach for delay sensitive and data reliable applications.

SVD-BASED SIGNATURE VERIFICATION TECHNIQUE USING DATA GLOVE

Data glove is a new dimension in the field of virtual reality environments, initially designed to satisfy the stringent requirements of modern motion capture and animation professionals. In this paper, we try to shift the implementation of data glove from motion animation towards signature verification problem, making use of the offered multiple degrees of freedom for each finger and for the hand as well. The proposed technique is based on the Singular Value Decomposition (SVD) in finding r singular vectors sensing the maximal energy of glove data matrix A, called principal subspace, and thus account for most of the variation in the original data, so the effective dimensionality of the data can be reduced. Having identified data glove signature through its rth principal subspace, the authenticity can then be obtained by calculating the angles between the different subspaces. The SVD-signature verification technique is tested with large number of authentic and forged signatures, showing remarkable level of accuracy in finding the similarities between genuine samples as well as those differentiated between genuine-forgery trials.

Epileptic seizure detection using the singular values of EEG signals

A new technique based on Singular Value Decomposition (SVD) for the detection of epileptic seizures is proposed. The SVD is applied sequentially on a sliding window of one second width of EEG data and the r singular values are obtained and used to indicate sudden changes in the signals. EEG recordings of 4-paediatric patients with 20 seizures are used to validate the proposed algorithm and the preliminary results indicates good level of sensitivity by the singular values to the changes in the EEG signals due to epileptic seizure. This sensitivity can be used to develop more reliable seizure detector than the existing techniques.

Multilevel thresholding for segmentation of pigmented skin lesions

For automatic detection of skin lesion, segmentation has always been a primary issue. Many techniques for segmentation and border detection of skin lesion have been discussed in literature. In this paper, we have proposed a multilevel thresholding algorithm which is primarily based on Otsus method to find between class variance of image histogram. Our algorithm is able to divide image into multiple classes based on intensity level with optimized selection of threshold values. Our proposed approach has shown quite reasonable results and is computationally fast and easy.