Розробка алгоритму виявлення рухомих об’єктів в системах технічного зору

Abstract

This article presents a method for detecting moving objects in video sequences, based on the analysis of contrast differences between the current frame and a pre-formed background. The algorithm evaluates intensity variations in image regions relative to their corresponding background values, allowing it to identify areas where the contrast exceeds a specified threshold. These regions are then used to generate a binary mask representing potential moving objects, with subsequent steps including the removal of small noise components caused by lighting fluctuations or sensor imperfections and the grouping of continuous regions to form coherent object boundaries. The method is capable of reliably extracting objects with sufficient contrast, while low-contrast or very small objects may remain undetected, and its simplicity and low computational cost make it suitable for real-time applications such as video surveillance, automated monitoring of static scenes, and security systems in IoT or embedded platforms.