Autonomous obstacle avoidance and navigation method for unmanned aerial vehicle-based field search and rescue

Abstract

An autonomous obstacle avoidance and navigation method for unmanned aerial vehicle (UAV)-based field search and rescue provided by the invention includes global navigation based on GPS signals and anautonomous obstacle avoidance and navigation algorithm based on binocular vision and ultrasound. First, a UAV carries out search under the navigation of GPS signals according to a planned path. Aftera target is found, the autonomous obstacle avoidance and navigation algorithm based on binocular vision and ultrasound is adopted. A binocular camera acquires a disparity map containing depth information. The disparity map is re-projected to a 3D space to get a point cloud. A cost map is established according to the point cloud. The path is re-planned using the cost map to avoid obstacles. When there is no GPS signal, a visual inertial odometry (VIO) is used to ensure that the UAV flies according to the planned path, and an ultrasonic sensor is used to detect and avoid obstacles below the UAVto approach the rescued target. UAV autonomous obstacle avoidance and navigation in a field search and rescue scene is realized.

Description

technical field [0001] The invention belongs to the field of obstacle avoidance of unmanned aerial vehicles, and in particular relates to an autonomous obstacle avoidance and navigation method for unmanned aerial vehicle search and rescue in the wild. Background technique [0002] In recent years, with the gradual maturity of UAV control technology, its applications have become more and more extensive, including agricultural assistance operations, industrial power line inspections, etc., which involve many aspects of technology, such as target detection and tracking, aircraft movement control, long-distance signal transmission, etc. Autonomous obstacle avoidance and navigation is also an important part of UAV technology. It means that the UAV actively senses the surrounding obstacles through the sensor, and performs new path planning according to the obtained obstacle information to avoid the surrounding obstacles and move. Arriving at the target position is the basis for e...

AI Technical Summary

Problems solved by technology

Earlier, there was a plan to use multiple ultrasonic sensors for obstacle avoidance, but in the wild environment, obstacles are dense, and only ultrasonic sensors cannot obtain complete obstacle information, and cannot detect slender obstacles such as branches and wires, and cannot guarantee real-time updated so that moving obstacles cannot be detected

At present, in the field environment, there is no better algorithm that can obtain rich obstacle information, detect slender obstacles, and perform real-time update and path planning to ensure that the UAV can successfully complete the field search and rescue mission.

Images