Binocular vision-based unmanned aerial vehicle obstacle avoidance method and system

Abstract

The present invention relates to the unmanned aerial vehicle technology field, in particular to a binocular vision-based unmanned aerial vehicle obstacle avoidance method and system. The method comprises the following steps of obtaining an environmental image in front of an unmanned aerial vehicle via a binocular camera; according to the obtained environmental image, identifying whether a target object is in front of the unmanned aerial vehicle; if the target object exists, based on the current attitude information of the unmanned aerial vehicle and a depth value of the target object obtainedaccording to the environmental image, determining whether an obstacle exists in the front target object; if the obstacle exists in the target object, based on a target depth value, obtaining the distance between the unmanned aerial vehicle and the obstacle, and avoiding the obstacle according to the distance between the unmanned aerial vehicle and the obstacle; if the target object is just the ground, the unmanned aerial vehicle not adopting an obstacle avoidance measure. The binocular vision-based unmanned aerial vehicle obstacle avoidance method and the binocular vision-based unmanned aerialvehicle obstacle avoidance system provided by the present invention realize the function of effectively distinguishing the obstacle and the ground in a low-altitude flight environment, and avoid thephenomenon of taking the ground as the obstacle mistakenly during the flight to frequently hover.

Description

technical field [0001] The invention relates to the technical field of drones, in particular to a binocular vision-based obstacle avoidance method and system for drones. Background technique [0002] A drone is an unmanned aerial vehicle operated by a radio remote control device or its own program control device. With the development of UAV-related technologies and the complex changes in its application scenarios, higher requirements are put forward for its environmental perception capabilities. Among them, the ability of UAVs to avoid obstacles on the traveling route is very important. [0003] At present, most systems for UAVs to avoid obstacles use binocular vision for obstacle recognition. Specifically, a binocular camera is installed on the UAV to obtain the first image and the second image of the obstacle , and then according to the parallax principle and the generated stereo vision information, the three-dimensional space coordinates of the obstacle can be directly r...

AI Technical Summary

Problems solved by technology

[0004] When the multi-rotor UAV is flying forward, it needs to tilt forward and downward to make the rotor tension generate a horizontal component, so as to realize the forward flight movement. The binocular camera mounted on the front of the drone will detect the ground, misjudge the ground as an obstacle, and cause unnecessary hovering

Images