Alexei Belotserkovsky

• PhDr
• Head of Department at United Institute of Informatics Problems

This article formalizes the problem of semi-automatic construction of the contour of area objects from satellite multispectral images and presents a solution algorithm using PCA and Dijkstra's algorithm. The emphasis of the article is on optimizing PCA using annual matrices for images of the same class, as well as optimizing the Dijkstra algorithm...

This article describes a realtime algorithm to determine a person’s posture at a certain point in time.

Nowadays autonomous remote monitoring and control systems are required for the reliable operation of complex, especially distributed over large areas or inaccessible places, technological facilities and systems for various purposes (agriculture, urban studies, environmental protection, emergency natural and man-made situations, etc.), as well as fo...

In this paper a principle scheme of new approach to Reliable Unmanned Aerial Vehicles Navigation (UAV) is purposed. It consists in building a reliable reserve control channel and using it together with well-known GPS/INS scheme. The idea of the approach lies in the usage of information from the onboard camera directed at nadir to enhance positionin...

In this paper method of biological tissue image analysis is proposed. This method is based on scale properties of biological objects. Different characteristics of tissue are calculated on levels of scale image pyramid and allow defining more complete description for diagnostics or monitoring of disease.

In this paper an elegant concept is proposed, which is for process management in applied analytical systems. Described schemes are currently used for image processing, and allow designing adaptive graphic user interface based on intelligent social (recommending) component so to build-up a functionality with flexible scripts and use objective and re...

Conceptual schemes for process management of image processing system is proposed here so to generate a Graphic User Interface, according to specific needs of certain users and developers. Schemes allow modifications and building-up the functionality since it based on a script-kernel and intelligent support by collaborating approach, which gives abi...

A new methodology of the development of intelligent software for medical image analysis is proposed. The methodology is based on combining an intelligent agent and an interpreter. It works as a script generator for creating new sequences of image processing functions. The script generator forms new image processing scripts after analysis of imag...

Algorithm of magnetic structure extraction and analysis in ultra thin magnetic film images is proposed. This algorithm allows to gain successful results and calculates a large number of characteristics of structures on magneto-optical images. The algorithm consists of three parts: background correction, structure extraction and calculation of dendr...

Описывается алгоритм сегментации компьютерно-томографических изображений, в основе которого лежит анализ глобально-локальной информации изображения. Реализовано дополнительное представление, в котором локальные свойства выражены сгенерированными псевдоизображениями из первой и второй производных оригинала. Вместе они составляют трехмерное пространс...

The diagnostics of oncological diseases is based on histological specimens in hematoxilin-eosin staining. Since manual evaluation of microscopy images is time consuming and depends on the human expert, several approaches to automatic image analysis and classification have been published. In such systems, feature extraction usually relies on a fixed...

The most interesting Belarussian medical systems which deal with histological images to carry out an analysis of histological and cell structures of different tissues are presented here. In this paper basic properties of system "Bioscan-IW", "AutoScan", "Contour" and "Cytron" are described. These systems are the most interesting among Belarussian h...

The quantitative measurement is basic part of histological cancer diagnostics. Measurement of orientation from tissues images allows to determinate symptoms of diseases. In this paper the algorithm for determination of cells orientation is proposed. This algorithm consists of three stages: segmentation of cells, measuring and classification. .

The principal scheme for automation of preliminary histological diagnostics of oncological diseases is described in the paper. A brief review of systems which are close to this scheme and based on automated and semi-automated segmentation approaches and morphometrical analysis are presented here. Processing of images with different optical magnific...