Розпізнавання 2D об’єктів у потоці відеоданих

Abstract

The masters degree work proposes information technology for recognizing 2D objects in a video data stream. The basic methods of image recognition were determined, their advantages and disadvantages were determined, methods of determining the key points in the image and methods of describing such key points were considered. Based on this, a choice was made in favor of the SIFT recognition technique. Stages of information technology for recognizing 2D objects in the stream of video data, as well as its program implementation in the Python programming language in the PyCharm environment using the specialized OpenCV library, were developed. The software is characterized by an elevated analogue value of 7.2% of the reliability of the recognition of 2D objects in the stream of video data.