RUS  ENG
Full version
JOURNALS // Vestnik Sankt-Peterburgskogo Universiteta. Seriya 10. Prikladnaya Matematika. Informatika. Protsessy Upravleniya // Archive

Vestnik S.-Petersburg Univ. Ser. 10. Prikl. Mat. Inform. Prots. Upr., 2015 Issue 2, Pages 115–122 (Mi vspui247)

This article is cited in 1 paper

Computer science

Digital images processing based on graph stationary flow construction

A. M. Batyukov

St. Petersburg State University, 7/9, Universitetskaya embankment, St. Petersburg, 199034, Russian Federation

Abstract: A modification of the algorithm to construct a stationary flow on a graph, which is applied to classification of digital images, is considered. An adaptation of the algorithm for multi core systems by the image partitioning is discussed. The dependence of calculation rate and classification accuracy on the number of partition elements is investigated. The numerical results are given. Bibliogr. 7. Table 2.

Keywords: image analysis, stationary flow on a graph, maximization of weighted entropy, parallel programming.

UDC: 004.942

Received: February 17, 2015



Bibliographic databases:


© Steklov Math. Inst. of RAS, 2026