基于LSD检测的无人机视觉着陆定位算法

Visual Landing Algorithm of Fixed-wing UAV Based on LSD Detection Algorithm

摘要:

针对复杂电磁和拒止环境下固定翼无人机自主着陆应用场景,提出了一种基于LSD(Line Segment Detector)的飞机着陆跑道检测和定位的方法；根据机场数据特点,构建机场跑道数据模型,并对数据进行标定和预处理,采用LSD直线检测算法以提高检测时间,采用蒙版技术以提高跑道的检测精度,通过检测得到机场跑道线在图像坐标系下的信息；根据相机成像原理推导出基于线检测的PNP(Perspective-n-Points)定位算法,从而求出飞机相对于跑道的位置；分别在视景仿真环境和真实机场环境进行检测和定位解算,结果表明本文设计的算法满足无人机着陆定位的实时性要求。

Abstract:

Aiming at the application scenario of autonomous landing of fixed wing unmanned aerial vehicles in complex electromagnetic and GPS-denied Environments, a method for detecting and locating aircraft landing runways based on LSD (Line Segment Detector) is proposed; According to the characteristics of airport data, an airport runway data model is constructed, and the data is calibrated and preprocessed. LSD line detection algorithm is used to improve detection time, and masking technology is used to improve the detection accuracy of the runway. Through detection, the information of the airport runway line in the image coordinate system is obtained; According to the camera imaging principle, a PNP (Perspective n Points) positioning algorithm based on line detection is derived to calculate the position of the aircraft relative to the runway; The results of detection and location calculation in visual simulation environment and real airport environment show that the algorithm designed in this paper meets the real-time requirements for landing location of unmanned aerial vehicles.