基于图像识别的无人机导航系统设计  

Design of UAV Navigation System Based on Image Recognition

作　　者：李高杰[1] Li Gaojie(Zhengzhou Institute of Finance and Economics,Zhengzhou 450000,China)

机构地区：[1]郑州财经学院,郑州450000

出　　处：《农机化研究》2021年第1期114-118,共5页Journal of Agricultural Mechanization Research

基　　金：河南省高等学校重点科研项目(18B630016)。

摘　　要：无人机提高了农业生产效率,但作业性能受各方面因素的影响,而导航能力是重点关注的内容。图像识别技术通过分析图像中的大量信息来辨识目标,可以提高无人机的自动导航能力。固定翼无人机在飞行速度和高度上具有优势,进行农业信息监测和保险勘察时效率更高,但精准导航的难度较大。为此,设计了一个基于图像识别的无人机导航系统,以固定翼的航拍无人机为平台,搭载影像传感器拍摄图像,进行图像边缘检测和目标识别分析,获得空间定位和跑道识别的功能。同时,地面站根据分析结果生成飞行控制指令,无人机接收指令后通过舵面偏转调整飞行姿态,获得自主飞行和着陆能力。UAV has improved the efficiency of agricultural production,and its performance is affected by various factors.Navigation ability is the focus of attention.Image recognition technology can improve the automatic navigation capability of drones by analyzing a large amount of information in the image to identify the target.Fixed-wing UAVs have advantages in flight speed and altitude,and are more efficient in agricultural information monitoring and insurance surveys,but precision navigation is more difficult.In this paper,a UAV navigation system based on image recognition is designed.The fixed-wing aerial drone is used as the platform,and the image sensor is used to capture images.The system performs image edge detection and target recognition analysis to obtain the functions of spatial positioning and runway recognition.The ground station generates flight control commands based on the analysis results,and the drone receives the command and adjusts the flight attitude through the deflection of the rudder surface to obtain the capability of autonomous flight and landing.

关 键 词：无人机 图像识别 导航 飞行姿态 自主着陆 农业 

分 类 号：S252[农业科学—农业机械化工程]