Research On Distance Measurement Technology For Binocular UAV Obstacle

Most GPS/IMU integrated navigation systems used by drones can control drone flight,but they can't find and identify space obstacle information.This is undoubtedly a hidden safety hazard in UAV flight.How to make UAVs have the ability to avoid obstacles and ensure the flight safety of drones is one of the major problems that need to be solved in the application of small and medium drones.In this paper,the binocular vision technology is introduced into the autonomous obstacle avoidance of drones,and the technology of binocular vision system integration,stereo matching,obstacle extraction,obstacle distance measurement and other technologies of UAVs are researched to build a system based on binocular vision.The aircraft obstacle recognition and distance measurement system provides technical support for drone autonomous obstacle avoidance.The main work and conclusions of this paper are as follows:(1)Binocular stereo vision system integration and calibration: This article uses two fixed cameras of the same model to form a binocular vision system,integrates it with a drone to construct a drone binocular vision system,and uses the Opencv method to implement the camera.Calibration,video frame interception and stereo correction,and experimental verification of the effectiveness of the integrated program and the feasibility of calibration and image preprocessing methods.(2)Binocular Stereo Matching: A stereo matching method combining Fast and Sift was constructed and programmed using multi-threaded programming,which greatly improved the image matching efficiency.The experimental results show that this method can meet the demand of unmanned aerial vehicle obstacle avoidance.(3)Obstacle extraction and distance measurement: In this paper,an automatic extraction method for obstacles using contour detection and image segmentation is presented.A better obstruction detection effect is achieved,and obstacles are completed using triangulation.Distance measurement.Experiments show that the proposed method can achieve obstacle distance measurement,and after adopting multi-threaded calculation strategy,it can meet the requirements of obstacle avoidance of the drone in terms of timeliness.(4)UAV binocular vision distance measurement and accuracy analysis were conducted in multiple indoor obstacles,outdoor buildings,trees,and multiple obstacle environments.The results show that the measurement accuracy of this method can be applied to autonomous obstacle avoidance in drone flight.