作　　者： ; ;

机构地区： 广东第二师范学院

出　　处： 《现代电子技术》 2018年第16期132-134,139,共4页

摘　　要： 基于鲁棒滤波的无人机着陆相对导航方法是通过设计的鲁棒高阶容积滤波相对导航滤波器实现无人机着陆导航,缺乏无人机导航图像预处理过程,导航准确性差。为此,提出基于智能视觉的无人机着陆导航方法,采用基于视觉的无人机导航特征提取方法,通过高、低帽滤波器增强无人机视频跑道图像,使用Ostu分割法对增强无人机视频跑道图像做二值化处理和去噪预处理,通过Hough变换提取预处理后无人机跑道图像中跑道的直线特征。依据获取的直线特征,进行无人机姿态角和位置运算以及换算后,获取无人机的姿态角和位置,实现无人机着陆的自主导航。实验证明,所提方法能够有效地对无人机着陆进行导航,并且准确性和效率较高。 The unmanned aerial vehicle(UAV)landing relative navigation method based on robust filtering achieves UAV landing navigation by means of the designed robust highorder volumefiltering relative navigation filter,resulting in lack of the preprocessing process of UAV navigation images,and poor navigation accuracy.Therefore,a UAV landing navigation method based on intelligent vision is proposed.The visionbased UAV navigation feature extraction method is adopted.The high and low hat filters are used to enhance UAV video runway images.The Ostu segmentation method is used to perform binarization processing and denoising preprocessing for enhanced UAV video runway images.The Hough transform is used to extract the linear feature of runways in preprocessed UAV runway images.According to the obtained linear feature,the attitude angle and position of the UAV are obtained after calculation and conversion of the attitude angle and position of the UAV,so as to realize autonomous navigation of UAV landing.The experimental results show that the proposed method can effectively navigate UAV landing,and has high accuracy and efficiency.

关 键 词： 智能视觉 无人机 着陆导航 图像预处理 特征提取 姿态角

领　　域： []