一种基于改进ORB特征匹配的无人机视觉导航方法

为了解决在全球导航卫星系统(Global Navigation Satellite System)拒止情况下无人机导航能力缺失等问题，提出了一种基于改进快速提取旋转描述子(Oriented FAST and Rotated Brief, ORB)图像特征匹配的无人机视觉导航方法。首先，为了实现无人机的绝对定位，提出了一种特征图像基准数据库构建方法；其次，为提取图像数据集的特征点，采用了一种结合尺度不变特征变换(Scale Invariant Feature Transform, SIFT)的尺度空间优化ORB特征提取算法；最后，为了将图像特征与图像基准数据库快速匹配并提高其匹配精度，提出了一种改进ORB特征匹配算法——ORB+GMS+PROSAC算法。通过在ArcGIS中分割图像构建基准数据库并进行实验分析，结果表明，基于ORB+GMS+PROSAC特征匹配算法性能显著提升，其中匹配准确率上升5.05%,匹配时间减少41.61%,明显优于其他传统特征匹配算法。

A UAV Visual Navigation Method Based on Improved ORB Feature Matching

To solve the lack of UAV navigation capability in the case of GNSS-denied,an improved method of UAV visual navigation based on ORB image feature matching is proposed.First,a method for constructing a feature image benchmark database is proposed to achieve the absolute position of UAVs.Secondly,an ORB feature extraction algorithm combined with SIFT scale-space optimization is adopted to extract the feature points of the image dataset.Finally,an improved ORB feature matching algorithm is proposed,which is named ORB+GMS+PROSAC algorithm,to quickly match the image features with the image reference database and improve the matching accuracy.The result shows that the performance of the feature matching algorithm based on ORB+GMS+PROSAC is significantly improved by segmenting images in ArcGIS to construct a benchmark database and conduct experimental analysis.Among them,the matching accuracy rate increases by 5.05