Showing papers in "WSEAS Transactions on Signal Processing archive in 2011"

Handwritten signature identification using basic concepts of graph theory

Abstract: Handwritten signature is being used in various applications on daily basis. The problem arises when someone decides to imitate our signature and steal our identity. Therefore, there is a need for adequate protection of signatures and a need for systems that can, with a great degree of certainty, identify who is the signatory. This paper presents previous work in the field of signature and writer identification to show the historical development of the idea and defines a new promising approach in handwritten signature identification based on some basic concepts of graph theory. This principle can be implemented on both on-line handwritten signature recognition systems and off-line handwritten signature recognition systems. Using graph norm for fast classification (filtration of potential users), followed by comparison of each signature graph concepts value against values stored in database, the system reports 94.25% identification accuracy.

Improving speech intelligibility in cochlear implants using acoustic models

Abstract: Cochlear implant (CI) is a prosthetic device that partially replaces the functions of the human ear via electrical stimulation. Cochlear implants are system and/or patient specific that mandates a simulation model prior to implantation. In the present work to improve the perceptual quality of the speech generated by a CI model, system specific parameters are analyzed by developing uniform bandwidth filterbank-based acoustic CI models, an auditory model-based CI system with frequency bands spacing similar to the critical-bands of an auditory system and Mel-frequency cepstral coefficients (MFCC) based analysis-synthesis system for cochlear implants. Acoustic CI simulations are generated for all the vowels of English language and words (easy and hard) from the Lexical Neighbourhood Test (LNT) and sentences from TIMIT database using waveform and feature extraction strategies. A closed-set listening test is conducted and a comparative study is made among the various acoustic CI models developed. The perceptual quality/speech intelligibility of the speech is rated in 5 point grading. It is observed that the acoustic CI simulation for sentences generated by critical-band-based CI system showed a mean opinion score of 4.1 as opposed to 3.1 for uniform bandwidth filters-based CI system.

Laser scanner technology for complex surveying structures

Abstract: Generally, when someone refers to architectural property he inclines to consider only that part of architecture and monuments belonging at remote epochs far since our days. However authors' opinion is that the diffusion and the spreading of the culture cannot leave out from the analysis, the study and the conservation also of architecture and all things realized in more recent times. However, the characteristics of the modern and contemporary architecture with respect to those precedents lead to the development of a definition for new approaches and adequate representation forms because of the presence both of materials and innovative technologies like the tubular or trellis structures, that show then different difficulty in the interpretation and definition of the acquired data. In such direction, the new digital technologies allow, from a part, a rationalization and rapidity of the relief operations, from the other they allow to create some new representations which can easily fit to the scholars and operators (architects, engineers, restorers, historians, etc) demands, or to be used to produce faithful copies through quick prototype techniques, but also, more simply, to give back enjoyable such information easily by town councils or web users. The developed and described, in this article, experiences have the aim of verify the potentialities of laser scanner in surveying of structures, for whom traditional techniques of relief could result disadvantageous in terms of realization's times, costs and precision. A particular attention has addressed to elaboration phase, data filtering and 3D modeling through the use of specific and opportune algorithms of best-fitting, useful for individualization and extraction of forms.

A scalable architecture for H.264/AVC variable block size motion estimation on FPGAs

Abstract: In this paper, we investigate the use of Field-Programmable Gate Arrays (FPGAs) in the design of a highly scalable Variable Block Size Motion Estimation architecture for the H.264/AVC video encoding standard. The scalability of the architecture allows one to incorporate the system into low cost single FPGA solutions for low-resolution video encoding applications as well as into high performance multi-FPGA solutions targeting high-resolution applications. To overcome the performance gap between FPGAs and Application Specific Integrated Circuits, our design minimizes the increase in memory bandwidth as the design scales. The core computing unit of the architecture is implemented on FPGAs and its performance is reported. It is shown that the computing unit is able to achieve 58 frames per second (fps) performance for 640×480 resolution VGA video while incurring only 4.5% LUT and 6.3% DFF utilization on a Xilinx XC5VLX330 FPGA. With 8 computing units at 38% LUT and 55% DFF utilization, the architecture is able to achieve 50 fps performance for encoding full 1920×1088 progressive HDTV video.

A fast zigzag-pruned 4×4 DTT algorithm for image compression

Abstract: The Discrete Tchebichef Transform (DTT) is a linear orthogonal transform which has higher energy compactness property like other orthogonal transform such as Discrete Cosine Transform (DCT). It is recently found applications in image analysis and compression. This paper proposes a new approach of fast zigzag pruning algorithm of 4×4 DTT coefficients. The principal idea of the proposed algorithm is to make use of the distributed arithmetic and symmetry property of 2-D DTT, which combines the similar terms of the pruned output. Normalization of each coefficient is done by merging the multiplication terms with the quantization matrix so as to reduce the computation. Equal number of zigzag pruned coefficients and block pruned coefficients are used for comparison to test the efficiency of our algorithm. Experimental method shows that our method is competitive with the block pruned method. Specifically for 3×3 block pruned case, our method provides lesser computational complexity and has higher peak signal to noise ratio (PSNR). The proposed method is implemented on a Xilinx XC2VP30 FPGA device to show its efficient use of hardware resources.

An effective joint implementation design of channel equalizer and DDC for WDAR receiver

Abstract: In this paper, the technique of digital down conversion (DDC) and channel equalization in wideband digital array radar (WDAR) are introduced. The analysis about the frequency domain equalization algorithm based on weighted least squares (WLS) is derived in detail at first. Then, an efficient DDC based on polyphase structure with intermediate frequency (IF) bandpass sampling is presented. Moreover, a simple structure which combines equalizer and DDC is proposed, which is proven to be valid, feasible and efficient. A design example is given and the FPGA resource consumption saving is discussed also. The corresponding simulations and test results demonstrate the effectiveness of the proposed DDC and show that the performance of the channel mismatch after equalization is improved obviously.

Biorthoganal wavelet packets and Mel scale analysis for automatic recognition of Arabic speech via radial basis functions

Abstract: In this paper, a Neural Network (NN) approach for the recognition of the Arabic digits is presented. The two phases of training and testing in a Radial Basis Functions (RBF) type network is described. Biorthogonal Wavelets are constructed and used for analysis of generated subwords of the digits. This approach decomposes spoken Arabic digits based on the acoustical information contained within the speech signals. The procedure locates the boundaries between subwords by finding the peaks in the function representing the spectral changes between consecutive speech frames. The Frame-based energy parameters derived from a Wavelet Packet Scale (WPS) are used in deriving the Spectral Variation Function (SVF). Three Biorthogonal wavelets are used as analyzing functions and their performances are compared with that of their Orthogonal counterpart and with that of the traditional Fourier based Mel scale approach.

A modified MIMO radar model based on robustness and gain analysis

Abstract: Multi-Input Multi-Output (MIMO) radar is a new radar technology in recent years, which partially or completely uses spatial diversity gain of the signal to replace coherent gain in traditional phased-array radar. Using the ideal point source model, we make a detailed analysis of the contributions to radar detection system made by these two kinds of gain. These contributions are divided into two kinds: the contribution to system robustness and the contribution to improving the signal-to-noise ratio. Based on this, it is proposed that the space diversity gain of MIMO radar can make more contribution to the system. The rationality of this proposal is further proved by the modification of the statistical MIMO model. And the theory above is verified by simulation. In addition, this paper illustrates how to analyze other MIMO radar systems from the viewpoints of these two kinds of contributions.

An adaptive stochastic-resonance-based detector and its application in watermark extraction

Abstract: In this paper, we explore a stochastic resonance (SR) based detector using bistable system (BS) to detect a binary pulse amplitude modulated (PAM) signal embedded in non-Gaussian noise. Through the example of BS based watermark extraction, we show that a reliable performance cannot be obtained if the BS parameters are determined by traditional tuning technique. The key observation is that the BS parameters are not sensitive to the pdf of the noise but to the variance of the noise and the amplitude of the signal. That makes it possible to determine the BS parameters in advance and an adaptive BS can be constructed based on the estimated amplitude of the watermark (signal) and the variance of the DCT coefficients (noise). Experimental results show that the performance obtained from the proposed adaptive stochastic-resonator-based detector is stable and provides superior performance compared to the existing BS based watermark schemes and the Gaussian based maximum likelihood (ML) detector.

Intercept of frequency agility signal using coding Nyquist folding receiver

Abstract: The parameter estimation of the frequency agility (FA) signal using the coding Nyquist folding receiver (CNYFR) is presented. The estimation algorithm adopting linear frequency modulation (LFM) as the local analogue modulation is derived. The Nyquist zone is estimated by the pseudo Wigner-Ville distribution (PWVD) and the hopping frequencies are calculated by the maximum likelihood (ML) method. Simulations show that CNYFR with analogue modulation of LFM has better performance than the sinusoidal frequency modulation (SFM) one, and the parameter estimation accuracy is acceptable when the SNR is above 0dB.

Improved beamspace MUSIC for finding directions of BPSK and QPSK coherent arrivals

Abstract: Coherence is often encountered in sonar, radar or mobile communications. Performance of the beamspace MUSIC (MUltiple Signal Classification) deteriorates as coherent arrivals become closely space. An improvement for the beamspace MUSIC resolution is presented. Decorrelation as a preprocessing is performed by the techniques called forward-backward spatial smoothing. Based on the fact that signal eigenvectors of the smoothed correlation matrix of the received signals contains the DOA vectors, combining all signal eigenvectors into a signal sequence enable us to obtain estimated DOAs. This combined signal eigenvector is equivalent to an array output impinged by the partially correlated sources. As a consequence, forwardbackward averaging is presented to decorrelate the coherence in the correlation matrix of the combined signal eigenvector before applying the beamspace MUSIC to extract the DOA information. Evaluations are given to illustrate the capability of the proposed method to distinguish closely spaced directions and reduce estimation errors in the presence of fully correlation. Performance analysis of BPSK and QPSK modulations are derived and compared with the simulation results.

The use of wavelet entropy in conjuction with neural network for Arabic vowels recognition

Abstract: In this research paper, Arabic vowels recognition system using very promising techniques; wavelet packet transform (WT) with entropy and neural network was presented. Trying to enhance the recognition process, three types of entropies were applied for the wavelet packet (WP) of the speech signals. Moreover, different levels of WP were used in order to enhance the efficiency of the proposed work until level 7. To classify among the feature vectors; a probabilistic neural network (PNN) were used. A MATLAB program was used to build the model of the proposed work to show the powerfulness of 96.77% identification rate. This is due to that the functions of features extraction and classifications are performed using the entropy, wavelet packet and neural networks.

Efficient wavelet-based scale invariant features matching

Abstract: Feature points' matching is a popular method in dealing with object recognition and image matching problems. However, variations of images, such as shift, rotation, and scaling, influence the matching correctness. Therefore, a feature point matching system with a distinctive and invariant feature point detector as well as robust description mechanism becomes the main challenge of this issue. We use discrete wavelet transform (DWT) and accumulated map to detect feature points which are local maximum points on the accumulated map. DWT calculation is efficient compared to that of Harris corner detection or Difference of Gaussian (DoG) proposed by Lowe. Besides, feature points detected by DWT are located more evenly on texture area unlike those detected by Harris' which are clustered on corners. To be scale invariant, the dominate scale (DS) is determined for each feature point. According to the DS of a feature point, an appropriate size of region centered at this feature point is transformed to log-polar coordinate system to improve the rotation and scale invariance. To enhance time efficiency and illumination robustness, we modify the contrast-based descriptors (CCH) proposed by Huang et al. Finally, in matching stage, a geometry constraint is used to improve the matching accuracy. Compared with existing methods, the proposed algorithm has better performance especially in scale invariance and blurring robustness.

Intelligent infrared target for training commandos to combat urban terrorism

Abstract: The commandos are required to possess very quick reflexes for a surprise enemy presence in combing operations particularly in terrorist's attacks in urban areas. Therefore, it is necessary for commandos to receive rigorous training in environments that simulate real-life urban combat conditions as closely as possible. In this paper, we designed an Infrared (IR) target for commando training in urban areas. The designed system is based on Infrared technology which will train the commando in a simulated environment and evaluates the performance based on his response time in a given short period. The system has Passive Infrared Sensor (PIR) and IR Transmitter combined in one module and a microcontroller based target.