Implementation of Vision Based UAV Positioning System

Abstract

Since the emergence of Unmanned Aerial Vehicle (UAV), UAV has played an irreplaceable role in various fields with its unique advantages, such as flexibility, easy manipulation and low cost. Due to the low navigation accuracy and reliability of traditional inertial navigation, it has gradually been unable to meet the needs of people to perform high-precision flight missions. Moreover, the complex unfamiliar environment and electronic interference pose new challenges to the traditional GPS positioning system. This makes the UAV need to perceive the surrounding environment and make independent decisions with its own sensors in the face of complex and changeable environment, so as to realize accurate positioning and autonomous landing without traditional GPS signals. This paper proposes a solution based on the improved optical flow algorithm. The improved feature point matching algorithm can effectively improve the interference ability of the traditional optical flow algorithm in the face of illumination change noise, and the image segmentation algorithm is used to eliminate the interference of foreground motion noise, To a certain extent, it improves the accurate positioning ability of UAV in the face of complex and changeable environment and no GPS signal, and better improves the real-time performance of the algorithm.