International Journal of Tomography & Statistics ISSN 0972-9976 (Print); ISSN 0973-7294 (Online)
IJTS	Editorial Board		Call for Papers	Abstracts	Contents
Summer Special Volume Abstract
Volume 9		No. S08		Summer 2008

A Novel Up-To-Down Algorithm for Road Extraction David Tien^1, Yi Xiao^1and Qing-hua Qin^2 ^1School of Information Technology The University of Charles Sturt, Bathurst, NSW 2795, Australia ^2Department of Engineering the Australian National University, Canberra, ACT 0200, Australia ABSTRACT On the basis of the symmetric axis transform (SAT), the constrained Delaunay triangulation (CDT) technique, and a split-and-merge approach, a new ‘up-to-down’ algorithm is developed to identify various roads from aerial images. The contours (shapes) of all the potential road regions that are segmented from aerial images are represented by their SAT for decomposing these regions into parts, so that the linear features of the shapes such as the length, width and curvature can be quantitatively measured on the parts, while CDT technique is applied to implement the SAT in discrete domain. From the CDT, the length and width for each part can be computed, and a split-and-merge algorithm is applied for calculating the curvature of each part. Three rules are proposed in terms of the width, length and curvature to identify roads from the candidature road regions. The study shows that the proposed technique is a promising algorithm for identifying roads from aerial images and appears to be superior to the existing methods in that it can extract road networks under complex background with less artifacts. Moreover, as the symmetry based method can quantitatively describe roads, it is also a useful tool for road vectorization. Keywords: road extraction, symmetry axis transform, curvature calculation. 2000 Mathematics Subject Classification: 68U10, 62M40, 62H35.
Stable Gradient–Type Iterative Methods for Smooth Irregular Operator Equations and their Application to the Problem of Acoustic Sounding Mihail Yu. Kokurin^1 and Alexander I. Kozlov^2 ^1Department of Mathematics Mary State University, Lenin sqr. 1, Yoshkar–Ola 424001, Russia ^2Department of Mathematics Mary State Pedagogical Institute, Kommunisticheskaya 44, Yoshkar–Ola 424002, Russia ABSTRACT In this paper we present a general scheme for constructing stable iterative methods to solve nonlinear irregular equations with smooth operators in a Hilbert space. The approach is based on restricting Tikhonov’s functional with a nonnegative regularization parameter to an appropriate finite–dimensional affine subspace. In the subspace we search for a domain where the functional is strongly convex and has an unconstrained local minimizer. The minimizer serves as an approximation to a solution of the original equation. The application of relaxational iterative processes to local finite–dimensional minimization of Tikhonov’s functional generates a class of stable iterative methods for nonlinear irregular equations with arbitrary smooth operators. The suggested scheme is demonstrated by solving a model 3D problem of acoustic sounding. Keywords: irregular equations, iterative methods, stable iterations, inverse problems, acoustic scattering. 2000 Mathematics Subject Classification: 47J06, 49M20, 65J20, 74J25.
Classification of Soil Texture Based on Wavelet Domain Singular Values S. Ramakrishnan and S. Selvan Department of Information Technology, PSG College of Technology, Coimbatore-641 004, India ABSTRACT Singular Value Decomposition (SVD) based novel approach using wavelet packet transformation is proposed for classification of soil textures. A procedure for classifying the textures in the presence of additive white Gaussian noise is introduced and this procedure is experimentally validated. The proposed approach extracts features such as energy, entropy, local homogeneity and min-max ratio from the singular values of wavelet packet transformation coefficients. A modified form of Probabilistic Neural Network (PNN) called Weighted PNN (WPNN) is employed for performing the classification. Compared to probabilistic neural network, WPNN includes weighting factors between pattern layer and summation layer of the PNN. Experiments have been carried out to test the performance of the proposed approach in terms of classification rate at various Peak Signal-to-Noise Ratio (PSNR), various number of training texture images, various levels of wavelet packet transformation, and various feature set dimensionality. Experimental results showed superiority of the proposed approach over wavelet domain Gray Level Co-Occurrence Matrix (GLCM) based approach, wavelet domain SVD model based approach and Hidden Markov Model(HMM) based approach. Index Terms: Image Classification, Wavelet Packet Transformation, Probabilistic Neural Network Mathematics Subject Classification: 62H30, 65T60, 68T10, 74E25
Adaptation of Generalizability Theory for Inter-Rater Reliability for Landmark Localization Ilker ERCAN,^1Gokhan OCAKOGLU,^1Ibrahim GUNEY, ^2 and Berna YAZICI^3 ^1 Uludag University, Faculty of Medicine, Department of Biostatistics, Bursa, TURKEY ^2 Istanbul Aydin University, Science Faculty, Department of Statistics, Istanbul, TURKEY ^3 Anadolu University, Science Faculty, Department of Statistics, Eskisehir, TURKEY ABSTRACT Our study on inter-rater reliability for landmark localization is designed to observe the consistency in locating landmarks of the same or different rater replication on two or three dimensional forms. In the study, the input for the calculation of inter-rater reliability for landmark localization is the Euclidean distance of all landmarks located by the raters. We calculated the reliability coefficient for two-facet crossed design (landmark pairs-by-rater-by-subject) based on the Generalizability Theory (GT). In GT, the score reliability for relative (norm-referenced) interpretations is referred to as the generalizability (G-) coefficient. Key Words: generalizability theory, inter-rater reliability, landmark localization, landmark reliability, shape analysis Mathematics Subject Classification: 97C40
Estimation of the Seismic Dispersion Parameters by Wavelets Youssef Bentaleb¹ and Saïd El Hajji² ¹Laboratory of Mathematics, Computing and Applications Faculty of Sciences, Rabat-Agdal; Morocco ²Laboratory of Mathematics, Computing and Applications Faculty of Sciences, Rabat-Agdal; Morocco ABSTRACT In this paper, we propose a new approach for the dispersion estimation of a seismic surface waves, in this issue, our method is based on the Continuous Wavelet Transform (CWT) applied to the seismic signal (1D analysis). In fact, the proposed method give the dispersion estimation in two cases: in the first, which the propagation is modeled by a time delay and phase shift of wave between sensors, in the second case, the propagation take place in the complex field (inhomogeneous medium) thus modeled by three parameters: delay, phase shift and dispersion coefficient. When applied to synthetic data, this new method gives much improved results when compared to other standard methods. Keywords: Continuous Wavelet Transform, surface seismic waves, mathematical modeling, delay, phase shift, dispersion coefficient. 2000 Mathematics Subject Classification: 42C15, 42C99.
An Enhanced Watershed Transformation Approach for MRI Gray Matter Segmentation Using Iterative Parallel Region Merging K. Santle camilus^1, V. K. Govindan^2, P. S. Sathidevi^3 ^1,2Department of Computer Science and Engineering ^3Department of Electronics and Communication Engineering National Institute of Technology Calicut, NIT Campus P.O., Calicut, India - 673 601 ABSTRACT Accurate segmentation of cortical gray matter is important for a study of central nervous system diseases. Slice-by-slice manual segmentation of the cortical gray matter is a tedious and time consuming process. Automatic or semiautomatic segmentation using computer make the tough job easier for the radiologist to analyze the cortical gray matter. Among the existing segmentation algorithms, watershed transformation has proved to be very useful and powerful tool for morphological image segmentation because of its moderate computational complexity and its ability to identify vital closed boundaries of a given image even if the image contrast is poor. However, it exhibits over segmentation when applied to segment magnetic resonance image cortical gray matter. This work attempts to overcome the problem of oversegmentation by make use of a pre-segmentation and postsegmentation processes. Fuzzy filtering is employed as the pre-segmentation process which reduces the additional formation of local minima due to noise in the segmentation stage. A post-segmentation process, iterative parallel region merging is proposed in this paper which unites over segmented regions and identifies the existing natural segments of the magnetic resonance image. Keywords: Gray Matter Segmentation, Watershed Transformation, Fuzzy Filter, Iterative Parallel Region Merging. 2000 Mathematics Subject Classification: 68U10.
Adaptive Selectivity Frame for Image Denoising M. El Aallaoui,^1,2 A. El Bouhtouri,^1 and A. Ayadi^2 ^1Laboratoire d’Ingénierie Mathématique (LINMA) Département de Mathématique et Informatique, Faculté des Sciences - BP 20, ElJadida-Morocco ^2 Laboratoire de télécommunications et traitement de l’information, école Nationale des Sciences Appliquées Avenue Abdelkrim El Khattabi BP 575 Marrakech-Morocco ABSTRACT Two-dimensional wavelet analysis and directional frames are efficient in the analysis and the decomposition of oriented features in images. However, since wavelets share the same angular selectivity, isotropic, directional and less-oriented features are processed under the same framework with the same number of coefficients. We propose here a solution to solve this issue. We develop an adaptive representation for all image elements, ranging from highly directional ones to fully isotropic ones, by decomposing them into a frame of directional wavelets with variable angular selectivity. In the particular context of denoising of images plagued by white noise, after usual thresholding of the wavelet coefficients, our adaptive representation compares favorably to wavelet-based, curvelets and fixed selectivity reconstructions. Keywords: 2-D continuous wavelet transform; half-continuous directional frame; multiselectivity; adaptive selectivity; image denoising. 2000 Mathematics Subject Classification: 06D22, 42C40, 60G35, 94A08.
Analysis of Continuous-Time LMS Adaptive Filter Weights Using Stochastic Calculus and Fokker-Planck Kolmogorov Equation Tarun Kumar Rawat and Harish Parthasarathy Division of Electronics and Communication Engineering Netaji Subhas Institute of Technology, Sec-3, Dwarka, New Delhi ABSTRACT The LMS technique uses a simple approximation to the gradient to update it’s filter weight. This approximation, known as the noisy gradient, introduces a jitter into the LMS weight adaptation process and this jitter is present even at convergence. Consequently, the continuoustime LMS filter weights are stochastic processes, having time varying probability density function during the adaptation phase and a stationary probability density function after the filter has converged. In this paper, the probability density function of the continuous-time LMS adaptive filter weights is obtained. The LMS weight update is formulated as a stochastic differential equation for the system identification problem and the weight probability density function is next derived using a partial differential equation known as the Fokker-Planck Kolmogorov equation. Closed form solution is obtained for the steady state probability density function for the LMS weights. Mean and variance is also obtained in closed form directly from the stochastic differential equation for the LMS weights. Keywords: LMS adaptive filter, stochastic differential equation, Fokker-Planck Kolmogorov equation, probability density function, system identification. 2000 Mathematics Subject Classification: 93E24, 60H10, 60H30, 93E12.
IJTS	Editorial Board	Call for Papers	Abstracts	Contents