The multi-contrast MRI volume registration problem constitutes one of the basic steps in many computer vision applications. Since different types of MRI images have totally different intensity distributions, the use of well known area based image alignment techniques does not consist a good choice for the solution of the problem at hand. In this paper the usage of Self Quotient Volumes as an efficient representation of the candidate volumes, to permit the exploitation of area based techniques, is proposed. In addition, a well known "signal" similarity measure based on the Enhanced Correlation Coefficient, is adapted for solving the volume alignment problem. From the application of the proposed technique in a series of experiments with artificial, as well as real, MRI brain sequences and its comparison against state of the art volume registration techniques it seems that its performance is better for both types (same and different contrast) of alignment problem.

Repeated patterns (such as windows, balconies, and doors) are prominent and significant features in urban scenes. Therefore, detection of these repeated patterns becomes very important for city scene analysis. This paper attacks the problem of repeated pattern detection in a precise, efficient and automatic way, by combining traditional feature extraction with a Kronecker product based low-rank model. We introduced novel algorithms that extract repeated patterns from rectified images with solid theoretical support. Our method is tailored for 2D images of building façades and tested on a large set of façade images.

The performance of most of the image alignment algorithms degrades in the presence of untextured and homogeneous image regions. Such large regions, and thus inappropriate for using them for alignment, appeared in most of the biological objects. In this paper the Pixel-ECC image alignment algorithm to laparoscopic surgery tailored to avoid this form of degradation is proposed. This is achieved by the use of the Frobenius norms of the Hessian matrices of the image pair for the dynamic detection of the problematic regions in each iteration of alignment algorithm. The proposed algorithm as well as other state of the art image alignment algorithms are used in a number of experiments based on artificial and real laparoscopic data, and the proposed algorithm seems to outperform its rivals.

The accurate eye center localization constitutes a challenging problem. In this paper a new technique based on a modified version of Radial Symmetry Transform is proposed. Specifically, computing the components of the transform from different modalities of the original image, we succeeded to improve the accuracy of the eye center localizer. The results we obtained from the application of the proposed method to the most challenging face databases show that our technique outperforms in accuracy most of the state of the art techniques.

The automatic seismic signal detection constitutes a very interesting and challenging task. The main difficulty in solving this problem is attributed to the fact that both the statistical properties of seismic noise and the characteristics of the recorded events are in general unknown. In this paper, by exploiting the particular nature of the signals we are treating, as well as a number of very interesting properties possessed by exchangeable random variables, we formulate the problem at hand as a record segmentation one and propose the use of two functionally linked test statistics for its efficient and robust solution, in a two-step procedure. By following this approach, we succeed not only in detecting a seismic wave arrival, but also in identifying the entire interval occupied by the signal, while minimizing the number of the required parameters. The performance of the proposed technique is confirmed by a series of experiments, both in synthetic and real seismic data sets.

Repeated patterns (such as windows, tiles, balconies and doors) are prominent and significant features in urban scenes. Therefore, detection of these repeated patterns becomes very important for city scene analysis. This paper attacks the problem of repeated patterns detection in a precise, efficient and automatic way, by combining traditional feature extraction followed by a Kronecker product low-rank modeling approach. Our method is tailored for 2D images of building facades. We have developed algorithms for automatic selection of a representative texture within facade images using vanishing points and Harris corners. After rectifying the input images, we describe novel algorithms that extract repeated patterns by using Kronecker product based modeling that is based on a solid theoretical foundation. Our approach is unique and has not ever been used for facade analysis. We have tested our algorithms in a large set of images.

In this paper we propose a new iterative graph based clustering technique. The proposed method has three desirable characteristics as compared to well known clustering techniques. Specifically, because of its iterative nature, the number of clusters contained in a given data set it is not necessary to be known a priori as partitional clustering techniques demand. Its performance, in terms of the quality of the achieved clustering, does not depend on the distribution of the cluster's size. Finally, no threshold value is required for achieving the clustering as hierarchical clustering techniques demand. All these features manifest themselves in the important problem of clustering homologous proteins when only sequence information is available. The proposed method is tested against well known and widely used techniques, by conducting a number of experiments based on both artificial and real protein data sets and all above mentioned characteristics were confirmed.

The solution to the problem of seismic signals segmentation constitutes a very interesting and challenging task. The main difficulty in solving this problem is attributed to the fact that both the statistical properties of seismic noise, as well the characteristics of the recorded events are in general unknown. In this paper, by exploiting the particular nature of the signals we are treating, and by using some interesting properties that obeys a difference based test statistic as well as its ingredients, we propose an approach that results in a robust and efficient automatic detection method. From a series of experiments we have conducted in both synthetic and real seismic data, the effectiveness of the proposed technique is confirmed.

In this work we propose the use of a modified version of the correlation coefficient as a performance criterion for the image alignment problem. The proposed modification has the desirable characteristic of being invariant with respect to photometric distortions. Since the resulting similarity measure is a nonlinear function of the warp parameters, we develop two iterative schemes for its maximization, one based on the forward additive approach and the second on the inverse compositional method. As it is customary in iterative optimization, in each iteration, the nonlinear objective function is approximated by an alternative expression for which the corresponding optimization is simple. In our case we propose an efficient approximation that leads to a closed-form solution (per iteration) which is of low computational complexity, the latter property being particularly strong in our inverse version. The proposed schemes are tested against the Forward Additive Lucas-Kanade and the Simultaneous Inverse Compositional (SIC) algorithm through simulations. Under noisy conditions and photometric distortions, our forward version achieves more accurate alignments and exhibits faster convergence whereas our inverse version has similar performance as the SIC algorithm but at a lower computational complexity.

The ability of an algorithm to accurately estimate the parameters of the geometric transformation which aligns two image profiles even in the presence of photometric distortions can be considered as a basic requirement in many computer vision applications. Projective transformations constitute a general class which includes as special cases the affine, as well as the metric subclasses of transformations. In this paper the applicability of a recently proposed iterative algorithm, which uses the Enhanced Correlation Coefficient as a performance criterion, in the projective image registration problem is investigated. The main theoretical results concerning the iterative algorithm and an efficient approximation that leads to an optimal closed form solution (per iteration) are presented. Furthermore, the performance of the iterative algorithm in the presence of nonlinear photometric distortions is compared against the leading Lucas-Kanade algorithm by performing numerous simulations. In all cases the proposed algorithm outperforms the Lucas-Kanade algorithm in convergence speed and robustness against photometric distortions under ideal and noisy conditions.

The invariance of the similarity measure in photometric distortions as well as its capability in producing subpixel accuracy are two desired and often required features in most stereo vision applications. In this paper we propose a new correlation-based measure which incorporates both mentioned requirements. Specifically, by using an appropriate interpolation scheme in the candidate windows of the matching image, and using the classical zero mean normalized cross correlation function, we introduce a suitable measure. Although the proposed measure is a nonlinear function of the sub-pixel displacement parameter, its maximization results in a closed form solution, resulting in reduced complexity for its use in matching techniques. Application of the proposed measure in a number of benchmark stereo pair images reveals its superiority over existing correlation-based techniques used for subpixel accuracy.

FIR filters obtained with classical L/sub 2/ methods have performance that is very sensitive to the form of the ideal response selected for the transition region. This is because, usually, filter specifications do not constrain in any way the ideal frequency response inside this region. In this paper we propose a new general method for the weighted L/sub 2/ based design of arbitrary FIR filters. In particular we propose a well defined optimization criterion that depends on the selection of the desired response inside the transition regions. By optimizing our criterion we obtain desired responses that produce weighted mean square error optimum filters with extremely good characteristics. The proposed method is computationally simple since it requires the solution of a linear system of equations.

A well-known least squares optimum approximation method is proposed as an efficient initialization scheme for the Remez exchange algorithm. More specifically, we theoretically demonstrate that the "don't care" least squares optimum solution guarantees, inside the bands of interest, the correct number of alternating in-sign extrema of the error function, thus satisfying one of the two basic conditions that are sufficient for obtaining the L/sub /spl infin// optimum solution. Although convergence of Remez: is theoretically assured, its practical implementations may fail to converge in "difficult" design problems when classical initialization is used. In particular, Matlab's realization of Remez, when initialized with the proposed scheme, exhibits a significantly better overall performance that translates into faster convergence and more robust behavior, especially in difficult design problems.