OBJECT RECOGNITION AND TRACKING METHOD IN THE MOBILE ROBOT’S WORKSPACE IN REAL TIME

This article presents an object recognition and tracking method in the mobile robot’s workspace in real time. The main approach is to use a color mask, for which a mathematical description of the algorithm is proposed. The method is implemented in the Python programming language using the PyCharm development environment. During the research, experiments were carried out, on the basis of which important performance indicators were obtained. The processing time indicator, which measures the processing time of each frame of a video stream, demonstrated high efficiency, ranging from 0.0010 to 0.0020 seconds. Detection speed, which determines the speed of object detection in FPS, also presented good results, ranging from 501.47 to 1037.42 FPS.