Home – Home – Jornals – Avtometriya 2011 number 2

A new algorithm is proposed for estimating optimal threshold values in thresholding algorithms of wavelet filtration of signals and images. Numerous computational experiments are performed to compare this threshold with threshold values widely used in wavelet filtration algorithms. Application of the proposed threshold is demonstrated to ensure substantial reduction of the filtration error of both smooth and high-contrast images.

Specific features of operation of fallen person detection algorithms having a moderate computational complexity and based on a sequence of images obtained by stationary video surveillance cameras are analyzed. Approaches to enhancing the algorithm operation stability are proposed. One variant of obtaining a more reliable silhouette of the object is considered; parameters used in recognition of the object states are analyzed and selected. A method of increasing the detector stability is proposed. This method is based on identifying intervals of stability of the average values of parameters, which allows reliability of detection of the object states to be improved. Results of computational experiments are presented.

A pedestrian detection algorithm based on an analysis of the autocorrelation function of a seismic signal is proposed. The efficiency of this method in terms of detection and false alarm probabilities is shown. Approaches to reducing the complexity of the proposed algorithm are considered. A comparative study is made of the effects of the point sampling frequency and the method of calculating the autocorrelation function on the performance and complexity of the pedestrian detection algorithm.

This paper describes an optical measurement technique for the two-dimensional fringe pattern (by introducing the carrier frequencies in two spatial directions x and y) by the continuous wavelet transform (CWT) phase gradient method. Such transforms based on the Morlet and Paul wavelets have been applied to image rows and columns one by one and then added to find the final phase distribution, without using any unwrapping algorithms. This technique is compared with the S-transform phase gradient method. Numerical simulations and actual experiments are carried out to show the validity of this technique for finding the phase distributions.

The influence of the transportation lag and the classical methods of its correction in a control system for a mobile robot with a differential drive are considered. Such effects as slipping of wheels and surface roughness are demonstrated to restrict the area of applicability of conventional algorithms of time lag correction. A time lag correction method is proposed, which is based on analytical extrapolation and extended Kalman filter. Experimental results confirm the efficiency of the approach proposed.

An algorithm for obtaining invariants under a group of special affine transformations for vector fields defined by integral curves of dynamic systems. The algorithm for constructing topological invariants of vector fields is extended to the case where the elements of the special affine group differ at different points of an integral curve.

To explain the basic features of the response of disordered materials (glasses) to irradiation by light, a new approach based on the assumption that glasses have the so-called "medium-range" order (inhomogeneities with the characteristic scale of ~1 nm) is proposed and justified. The structure of inhomogeneities depends on the bond type: glasses with covalent directed bonds (strong glasses) at nanometer scales are similar to their crystalline ancestors. Glasses with non-directed bonds (ion and van der Waals bonds) do not form crystalline nanoinhomogeneities, though the existence of a certain medium-range order is typical for them. Using the concept of an inhomogeneous structure of glasses at nanometer scales, it is possible to explain the physical properties demonstrated by these materials: structural factor, specific features of the density of vibrational states in the terahertz range, mean free paths at frequencies below the boson peak frequency, optical memory phenomenon, and specific features of charge transfer. The proposed approach is an alternative for the widely used approach that involves construction of specific defects understood as violations of coordination of individual atoms, as quasi-molecular complexes, or as two-level states with no particular information about them. This approach reflects the specific features of the vitreous state, is universal, and is not based on a particular chemical nature of the material.

This paper considers features of the reflection of polarized light from a surface which moves relative to the observer with a velocity comparable to the velocity of light. The reflection coefficients and ellipsometric parameters of the moving object vary due to the transformation of the angle of incidence and wavelength of light, and depend on the orientation of the velocity vector with respect to the surface normal and the direction of the wave vector of the incident wave. Special cases of the mutual orientation of these vectors are presented. It is shown that if the velocity vector lies in the plane of the object and is perpendicular to the plane of light incidence, the isotropic reflecting surface exhibits anisotropic properties. An estimate is made of the limiting minimum velocity at which the anisotropy effect can be detected.

A theoretical and experimental study was performed of the interference of polarized e- and o-waves generated by passing a laser beam (collimated, convergent or divergent) through a variablesplitting-angle birefringent prism (VSABP) with inclined optical axes in the crystalline components of the prism. The VSABP-2 modification is suitable for measuring the radius of curvature of the incident wavefront from the deflection of interference fringes. This prism was used to develop an interferometric device for controlling the degree of divergence/convergence of laser radiation and a wavefront sensor for diagnosing flat wavefronts.

The use of laser Doppler anemometry methods for determining the radius and slip velocity of rolling wheels is justified theoretically and experimentally. It is demonstrated that these methods allow the wheel radius to be measured with a relative error of 0.05-0.02% for the wheel turning angle of 0.6-4.0º. The slip velocity can be calculated in this case with an error of 20 %. The slip velocity of a railroad wheel moving with a velocity of 60 mm/s is found.

This paper presents the results of a comparative analysis of high-speed fiber-optic communication lines based on phase and amplitude modulation formats