Autonomous Positioning and Control System of Rotor UAV Based on Machine Vision

Abstract

In recent years, with the rapid development of electronic technology and image recognition technology, target recognition technology based on rotary wing UAV has become a hot research topic. This paper mainly studies the autonomous positioning and control system of rotary-wing UAV based on machine vision. This paper uses the weighted average method to convert the color images collected by the camera into grayscale images. The color image collected by the airborne camera is grayed out, and the binary image is obtained after threshold segmentation. The median filter technology is used to eliminate the noise, the edge information of the mark is detected, and the Harris corner points are finally extracted. After the clustering operation is completed, the traditional least squares method is used to fit a straight line, each connected component is matched, and each point is weighted according to its gradient. In this paper, the RANSAC algorithm is used to remove the mismatch points and obtain the SIFT characteristic information. At the same time, the PID control algorithm is used to obtain the deviation required for PID control. According to the rotational speed of the four motors of the deviation control system, the attitude control of the aircraft is realized. Finally, the positioning accuracy of the system is evaluated. Experimental results show that the detection time of SIFT feature points is about 100ms. The results show that machine vision improves the positioning accuracy of rotary-wing UAV and improves the accuracy of target recognition.