Tara Singh Kamal

A new algorithm for skew detection and correction

Here we have proposed two algorithms. The first one detects the skewing of words and the second corrects the skewing from handwritten words. Both algorithms make use of the Radon transform based projection profile technique. The method does not require pre-processing and it works equally good even with noise. The method is fast. The algorithms have been tested on words taken from more than 200 writers and the results obtained confirm the overall accuracy of proposed system. No error was detected.

Development of ANN-based virtual fault detector for Wheatstone bridge-oriented transducers

This paper reports on the development of a new artificial neural network-based virtual fault detector (VFD) for detection and identification of faults in DAS-connected Wheatstone bridge-oriented transducers of a computer-based measurement system. Experimental results show that the implemented VFD is convenient for fusing intelligence into such systems in a user-interactive manner. The performance of the proposed VFD is examined experimentally to detect seven frequently occurring faults automatically in such transducers. The presented technique used an artificial neural network-based two-class pattern classification network with hard-limit perceptrons to fulfill the function of an efficient residual generator component of the proposed VFD. The proposed soft residual generator detects and identifies various transducer faults in collaboration with a virtual instrument software-based inbuilt algorithm. An example application is also presented to demonstrate the use of implemented VFD practically for detecting and diagnosing faults in a pressure transducer having semiconductor strain gauges connected in a Wheatstone bridge configuration. The results obtained in the example application with this strategy are promising.

ANALYSIS AND DESIGN OF CIRCULAR FRACTAL ANTENNA USING ARTIFICIAL NEURAL NETWORKS

A Neural Network is a simplifled mathematical model based on Biological Neural Network, which can be considered as an extension of conventional data processing technique. In this paper, an Artiflcial Neural Network (ANN) based simple approach is proposed as forward side for the design of a Circular Fractal Antenna (CFA) and analysis as reverse side of problem. Proposed antenna is simulated up to 2nd iteration using method of moment based IE3D software. Antenna is fabricated on Roger RT 5880 Duroid substrate (High frequency material) for validation of simulated, measured and ANN results. The main advantage of using ANN is that a properly trained neural network completely bypasses the complex iterative process for the design and analysis of this antenna. Results obtained by using artiflcial neural networks are in accordance with the simulated and measured results.

Design of Sierpinski Carpet Fractal Antenna using Artificial Neural Networks

This paper deals with utilization of artificial neural networks for the design of Sierpinski carpet fractal antenna. The difficulty in designing of fractal microstrip patch antennas is due to the involvement of large number of physical parameters and hence their associated optimal values. It is indeed very difficult to formulate an exact numerical solution through empirical studies based on practical observations. In order to circumvent this problem, an alternative solution is achieved using artificial neural networks. The proposed technique used feed-forward back-propagation artificial neural network (FFBP-ANN) with one hidden layer to approximate neural model of this antenna. Sierpinski carpet fractal antenna is simulated using IE3D software. The investigation is done between the ranges of frequencies from 1 to 20Ghzs. The results obtained by using artificial neural networks are in agreement with simulated results. KeywordsCarpet, Artificial Neural Networks, Sierpinski, Antenna.

Fusion for registration of medical images - a study

The paper is a study demonstrating the application of discrete multiwavelets in medical image registration. The idea is to improve the image content by fusing images like MRI, CT and SPECT images, so as to provide more information to the doctor. The process of fusion is not new but here the results of study have been compared with the results from FCM algorithm used for similar application. Multiwavelets have been used for better clustering, as their decomposition results were better than Daubechies decomposition. A new feature based fusion algorithm has been used. This method shows results better than other methods for image registration when the images have been taken for the same person at a particular angle. The selective fusion not only gives more information but also helps in disease detection.

Performance of coherent square MQAM with L/sup th/ order diversity in Rician fading environment

This paper deals with the theoretical symbol error rate (SER) analysis of coherent square M-ary quadrature amplitude modulation (MQAM) with L/sup th/ order diversity in frequency non-selective slow Rician fading environment corrupted by additive white Gaussian noise (AWGN). The diversity combining technique considered here is maximal ratio combiner (MRC). A closed form expression for SER of MQAM in Rician fading environment under MRC diversity reception is derived and analyzed. The derived expression is in terms of a single finite integral with an integrand composed of elementary (trigonometric and exponential) functions. Because of its simple form, this expression readily allow numerical computation for cases of practical interests. To examine the dependence of error rate performance of MQAM on Rician parameter K and number of diversity branches L the numerical results are plotted as average SER versus average signal to noise ratio (SNR) for various values of K and L. The solution presented is general enough so that it includes Rayleigh fading and non-fading as special cases. The results presented in this paper are expected to provide useful information for exploiting the use of diversity for improving the performance of radio systems over fading channels.

Performance of coherent square M-QAM with L/sup th/ order diversity in Nakagami-m fading environment

The symbol error rate (SER) performance of coherent square M-ary quadrature amplitude modulation (M-QAM) with L/sup th/ order diversity frequency non-selective slowly Nakagami-m (1960) fading environment corrupted by additive white Gaussian noise (AWGN) is presented. The diversity combining technique considered in this paper is maximal ratio combining (MRC) with identical channels. The derived expression for SER is in terms of a single finite integral with an integrand composed of elementary (trigonometric) functions. Because of its simple form, the expression readily allows numerical evaluation for cases of practical interests. The solution presented in this paper is general enough so that it includes half Gaussian fading (m=0.5), Rayleigh fading (m=1), and nonfading (m=/spl infin/) as special cases. The results are plotted as SER versus signal to noise ratio (SNR) for various values of m and L to examine the dependence of the performance of MQAM on m and L. The results presented are expected to provide useful information needed for exploiting the use of diversity for design of better communication systems in Nakagami-m fading environment.

BER performance of MQAM with L‐branch MRC diversity reception over correlated Nakagami‐m fading channels

Exact and closed form generalized expressions for bit error rate (BER) of M-ary quadrature amplitude modulation (MQAM) with L-branch maximal ratio combining (MRC) space diversity reception in fading channels are derived and analyzed. The fading channels are modeled as identical but correlated frequency-nonselective slow Nakagami-m fading channels corrupted by additive white Gaussian noise (AWGN). Analytical results obtained are in terms of few finite range integrals with an integrand composed of elementary functions. Because of their simple form, these analytical results readily allow numerical evaluation in cases of practical interest. The results are also general enough to include Nakagami-m fading channels with and without correlation, no diversity system, Rayleigh fading channels with and without correlation, and AWGN as special cases. The numerical results for the case of 16QAM are shown graphically and also in tabular form in order to examine the effects of fading severity, order of diversity, and branch correlation on the BER performance. The two correlation models considered are constant correlation model and exponential correlation model. One may be interested to know how the BER of MQAM is related to symbol error rate (SER) of MQAM. Therefore, the BER results obtained in this paper are also compared with that obtained directly from the SER. It is expected that the analytical results presented in this paper will provide a convenient tool for design and analysis of a radio communication system with space diversity reception in uncorrelated and correlated fading environment. Copyright

Neurocomputational Approach for Feed-Position Estimation in Circular Micro-strip Antenna

This paper presents a neurocomputational model for estimation of feed-position in circular microstrip antenna. The difficulty in computing the feed position in circular micro- strip antenna lies due to the involvement of a large number of physical parameters including their associated optimal values. It is indeed very difficult to formulate an exact numerical solution merely on practical observations based empirical studies. In order to circumvent this problem, an alternative solution is achieved using neurocomputational model. The proposed technique used feed-forward back-propagation artificial neural network (FFBP-ANN) trained with Levenberg-Marquardt algorithm. The results of neural estimation are quite promising.

ARTIFICIAL NEURAL NETWORK BASED SOFT ESTIMATOR FOR ESTIMATION OF TRANSDUCER STATIC NONLINEARITY

In this work, the development of an Artificial Neural Network (ANN) based soft estimator is reported for the estimation of static-nonlinearity associated with the transducers. Under the realm of ANN based transducer modeling, only two neural models have been suggested to estimate the static-nonlinearity associated with the transducers with quite successful results. The first existing model is based on the concept of a functional link artificial neural network (FLANN) trained with mu-LMS (Least Mean Squares) learning algorithm. The second one is based on the architecture of a single layer linear ANN trained with alpha-LMS learning algorithm. However, both these models suffer from the problem of slow convergence (learning). In order to circumvent this problem, it is proposed to synthesize the direct model of transducers using the concept of a Polynomial-ANN (polynomial artificial neural network) trained with Levenberg-Marquardt (LM) learning algorithm. The proposed Polynomial-ANN oriented transducer model is implemented based on the topology of a single-layer feed-forward back-propagation-ANN. The proposed neural modeling technique provided an extremely fast convergence speed with increased accuracy for the estimation of transducer static nonlinearity. The results of convergence are very stimulating with the LM learning algorithm.