Vol 7, No 2 (2008)

An extended account (including the bibliography) of the talk to commemorate the 70-th anniversary of professor I. N. Sisakian and mark the 20-th anniversary of the Image Processing Systems Institute of the RAS is presented. The narrative of milestones in professor I. N. Sisakian’s life and professional career (08.03.1938 – 09.11.1995) is followed by the review of his major scientific achievements. Advances made after his untimely death in the field of Computer Optics pioneered by professor I. N. Sisakian are also over viewed.

A magnetooptic effect which arises when the electromagnetic wave propagates through a two-layer structure made up of a metal diffraction grating and a magnetized dielectric layer is considered. The rigorous solution of the problem of diffraction by the structure suggests that there is a strong dependence between the intensity in the zero diffraction order and the layer magnetization. We offer a qualitative explanation that relates the said effect with waveguide modes.

The band structure of a photonic crystal with the clathrate Si-34 lattice comprising 34 lattice sites in the unit cell is studied. The lattice is found to have a large isotropic band gap for a wide range of permittivity. The relationship between the photonic band isotropy and the permittivity is deduced.

Light beams with the orbital angular momentum in the problem of optical trapping and rotation of micro-objects are studied. Diffractive optical elements that generate a set of concentric rings of equal intensity for the optical trapping and rotation of opaque micro-objects are considered.

Self-reproducing Gauss-Laguerre, Gauss-Hermite and Bessel mode beams and their possibility for reconstruction after obstacle (opaque or transparent with refractive index n = 1,5) distortion are considered in that paper. Optically trapped particle as an obstacle is regarded. A comparative analysis of Gauss-Laguerre, Gauss-Hermite and Bessel modes and superposition beams capacity for reconstruction is conducted.

In present work the algorithm of integumentary plants fabric visualization, and also superficial phytogenesis formations, is developed. As a base method was used method Monte-Carlo. Statistical tests were made in program environmental TracePro Expert. The received scheme allows to consider influence of surface structure and form on integrated and differential properties of backscattered optical radiation, at interaction with a plant leaf.

Experimental researches of cold work’s removal and formation recrystallizated structures conditions by laser heating have shown an opportunity of passing these processes in a range of recrystallization temperatures titanic low alloys. Researches properties of the processed materials have shown that laser anneal application for local softening before shaping of half-finished products from metal alloys allows to increase accuracy of manufacturing of details, to reduce springing and bending radiuses.

A model scheme for spectral visualization of biological media macrostructure is presented. Superficial tissue with heterogeneities was chosen as a basic investigation object. Realization of the task is performed with 3D Monte Carlo simulation of laser radiation propagation in multiple scattered media. Back scattering differential scheme is the technique in operation for numerical calculations results analysis. Used method adds up to compute a differential value of spectral backscattering density which may characterize a state of the irradiated tissue. In the paper we investigate the differential scheme opportunities. Results allow to determine differential backscattering intensity as the multiparametric function of media and heterogeneities features. The data are basic for diagnostic scheme software implementation.

Construction of a parallel algorithm in the cyclic counter-sweep method for solving band-type grid equations is discussed. A variant of the linear breakdown of the two-dimensional grid domain is considered. Comparison is made with the familiar algorithms, disadvantages and benefits of the algorithm under discussion are shown.

In this work, we describe a classical iterative Kachmazh method and a modification thereof using a relaxation parameter. We study at which rate the Kachmazh method with a relaxation parameter converges when applied to solving a least squares problem of large dimension. Recommendations are given for choosing the relaxation parameter in a particular case - when solving the problem of polynomial approximation of the experimental data in the least squares sense.

There considered detection methods of remote sensing spacecraft on board equipment abnormal functioning in real time, preventing methods of abnormal situation grow up to irreversible consequences, methods of automatic diagnosis of on board systems failures and their revival. There given results of autonomous analysis, diagnosis and further reconfiguration of spacecraft attitude position control loop.

We propose a steganalysis technique for the Invisible Personal Information algorithm intended for embedding digital watermarks (DWM) in polygraphic products and considered steganographically robust. The steganalysis technique in question involves two stages: first, out of the entire set of embedding keys a probable key subset is extracted and, then, the image is being prepared using each of the keys chosen so that a watermark can be discovered. Also, we investigate the robustness of the steganalysis algorithm relative to distortions due to printing and scanning of stegagrams with DWMs.

This work deals with method for localizing and estimating the iris characteristics in biometric eye images. A method is proposed for estimating boundaries by the second-order curves, which is based on a discrete Hough transform modification. An algorithm for iris localization in the diagnostic iris images is developed and implemented. The results of investigation of the iris characteristics estimation algorithm on natural diagnostic images are discussed.

This work is aimed at studying the efficiency of the features based on texture energy measures when solving the image segmentation problem. An algorithm for feature space generation based on the texture energy measures is proposed. The texture features are used for segmentation of diagnostic crystallogram images. Using test texture images, the segmentation quality is estimated and the feature efficiency is studied when solving the texture segmentation problem.

We discuss an approach to fundus image analysis that enables one to modify the diagnostic vessel parameters in such a way that the morphological peculiarities of tree-like structures are accounted for. A feature estimation method based on processing the geometric characteristics of central vessel line lobes is described. An approach relying upon partitioning the curves into lobes with use of the algorithm of visibility matrix construction is discussed. The approach developed makes it possible to construct a primary feature space that can be used for constructing morphological features invariant to various types of geometric distortions.

The system of monitoring and forecasting of extreme situations of natural, technogenic and biologo-social character for territories of the Samara area is developed. The mechanism of gathering, processing and an information transfer for its estimation and acceptance of administrative decisions is described.