无人机基于视觉自主着陆的跑道识别跟踪

跑道检测识别与跟踪是基于视觉的无人机自主着陆的前提和难点,本文根据固定翼无人机基于视觉自主着陆的特点设计了包括跑道检测、跑道特征提取、跑道识别和跑道跟踪的方案,并在ARM Cortex-A9处理器中基于Linux系统使用OpenCV实现了该方案.最后按照逐步递进的方式分别对检测、检测识别、检测识别与跟踪结果进行了实验验证,并对实时性进行了分析.实验结果表明,通过该方案可以准确地识别图像中的跑道并具有较快的跟踪速度.

Runway Recognition and Tracking Based on Autonomous Landing of UAV

The detection, recognition and tracking of the runway are the key points and promise of the vision-based UAV autonomous landing,in the paper, a scheme with considering the characteristics of the fixed wing UAV based on visual autonomous landing is proposed0, including the runway detection, the runway feature extraction, the runway recognition and tracking algorithms.The scheme is implemented based on the Linux system using OpenCV in the ARM Cortex-A9 processor.Finally, the detection,the detection and recognition, and the detection and tracking are experimentally verified in a step-by-step manner, and the real-time performance of the scheme is discussed.The experiment results show that the program can accurately identify the runway in the image and has a faster tracking speed.