THE ANALYSIS OF FUNCTIONAL OPTIMIZATION METHODS FOR VIDEO PROCESSING IN OPENCV

Abstract

The article is devoted to analysing modern video processing methods using the library of functions and algorithms for computer vision, image processing and general-purpose numerical algorithms OpenCV. The study analyses the library’s functionality and possible ways to optimise video content processing.The paper highlights the relevance of using OpenCV when working with video; in particular, the modular structure of the library, which includes the core, imgproc, video, highgui, features2d, calib3d and stitching modules, is considered. The primary attention in the paper is paid to optimisation methods, in particular, the use of image pyramids pyrUp and pyrDown, image blending using the addWeighted method, colour conversion to grayscale using the cv2.cvtColor method, fast movement through the frames of the source video file using the Seek method and trimming excess video fragments using the Clip method.The key results of the study are an analysis of the effectiveness of various functional optimisation methods, especially in the context of image processing, their mixing and creating transition effects between video frames, conversion to grayscale, navigation in the video stream without significant delays, as well as cropping unnecessary areas of video files. These methods have proven their effectiveness, increasing the speed of video processing.The study confirms the high effectiveness of using OpenCV for video processing and optimisation tasks. Implementing functional optimisation methods allows video editors to significantly improve their performance and reduce the load on computing resources.Further research can aim to integrate OpenCV with machine learning and artificial intelligence technologies for automated video processing. Such integration will allow for more intelligent and efficient video processing processes, opening up new opportunities for real-time and complex video effects.