Methods for Improving Visual Terrain Relative Navigation for Dynamic Aerial Systems

Abstract

One popular means of aerial localization and navigation in GPS-denied environments is visual terrain relative navigation. Terrain relative navigation involves performing image registration with sensed aerial camera imagery and georeferenced satellite maps to produce the geographic translation and rotation of the camera. One popular terrain relative navigation technique depends on matching feature descriptors. These features, however, are intolerant to major changes in perspective, light, vegetation, season, and other scene changes and produce excessive amounts of false matches. Alternatively, image correlation can be used for registering a sensed image to a reference image but is extremely intolerant to perspective differences for 6 degree of freedom camera systems. This research explores the use of a combination of corner detection and normalized cross correlation for aerial vehicles at different altitudes. New methods for using dynamic search windows within reference satellite imagery are explored to constrain the pose estimation and increase image matching accuracy. The algorithm is tested with both simulated aerial imagery and experimentally sensed imagery captured with rigid mounted cameras on unmanned aerial vehicles and high altitude balloons. It is evaluated on its successful match rate and pose estimation error compared to GPS. It has approximately 75% successful match rate in simulation and 20% successful match rate in experimental datasets. The filtered pose estimate error is decreased in simulation in effectively over 95% of the frames and 20% in the experimental cases. Integration of this algorithm with other navigational sensors and algorithms would provide improvements to the overall navigation solution.