Compensation for low light and contouring in image recognition systems

Abstract

The presented work is devoted to solving specific problems in information and measurement systems for pattern recognition. This works examines the problems associated with ensuring the operation of image recognition systems in conditions of insufficient illumination in the observed scene. This problems are most often expressed in the presence of low object-background contrast and lead to the appearance of a contour effect in video frames. These effects significantly limit the scope of application of image recognition systems. The work provides an analysis of typical causes of the problems considered, including technological limitations. An overview of methods to compensation such effects. In this works proposed our own methods that will partially compensate for the influence of the considered effects, capable of working on various classes of scenes. The proposed methods are based on image fusion methods. In order to expand the range of problems to be solved, the proposed methods use different mathematical apparatus and image processing methods with different performance. Theoretical justification of the proposed methods is presented. The proposed methods will compensate for the influence of the effects under consideration and, as a consequence, improve the quality of operation of industrial image recognition systems. Compensating for the influence of the considered effects using the proposed methods can serve as the basis for expanding the scope of application of image recognition systems.

Author Biographies

Dmitry Aleksandrovich Abramov, Tula State University

Associate Professor of the Chair of Information Security, Cand. Sci. (Eng.)

Vyacheslav Leonidovich Tokarev, Tula State University

Associate Professor of the Chair of Information Security, Dr. Sci. (Eng.), Professor

Published
2023-12-20
How to Cite
ABRAMOV, Dmitry Aleksandrovich; TOKAREV, Vyacheslav Leonidovich. Compensation for low light and contouring in image recognition systems. Modern Information Technologies and IT-Education, [S.l.], v. 19, n. 4, dec. 2023. ISSN 2411-1473. Available at: <http://sitito.cs.msu.ru/index.php/SITITO/article/view/991>. Date accessed: 13 sep. 2025.