摘要: This work describes an algorithm for GPS denied navigation which uses Al-based pattern matching with Feature-Augmented Terrain Referenced Navigation. sUAS-borne radial-scan LiDAR can be used to generate a local high-spatial-resolu... 展开 This work describes an algorithm for GPS denied navigation which uses Al-based pattern matching with Feature-Augmented Terrain Referenced Navigation. sUAS-borne radial-scan LiDAR can be used to generate a local high-spatial-resolution point-cloud (LPC) elevation map. This map is then used for onboard/real-time registration to a hybrid elevation map, that integrates high- resolution feature point clouds with lower-resolution terrain elevation maps. This enables geographic location and altitude of the sUAS as measurements for the navigation filter. GPS denied features are then incorporated into autonomous navigation framework. 收起