基于机器视觉的无人机自主着陆技术

为了提高无人机着陆过程中的自主性和智能性,提出了一种基于机器视觉的无人机自主着陆算法。算法采用了红外图像与可见光图像协同的方式,首先对着陆模型进行设计;其次,通过着陆模型的颜色、纹理、热成像等特征对着陆模型进行检测识别;最后,通过确定降落模型的质心位置并跟踪,实现无人机的位姿调整。实验表明,该算法大大降低了基于机器视觉进行无人机位置识别时对环境光线的要求,提高了基于机器视觉的无人机自主着陆控制系统的抗干扰能力,实现了基于机器视觉进行自主降落的无人机的全天候自主着陆。

Autonomous landing technology of UAV based on machine vision

In order to improve the autonomy and intelligence of UAV landing process,an autonomous landing algorithm based on machine vision is proposed.The algorithm combines the way of infrared image and visible image.Firstly,the landing model is designed;secondly,the landing model is identified by its color,texture,thermal imaging and other features;finally,by determining and tracking the center of mass position of the landing model,the position and attitude of the UAV can be adjusted.Experimental results show that the algorithm greatly reduces the requirements of environmental light for UAV position recognition based on machine vision,improves the anti-interference ability of UAV autonomous landing control system based on machine vision,and realizes all-weather autonomous landing of UAV Based on machine vision.