An Interpolating Wavelet Filter for Terrain Navigation
Niclas Bergman · 1998
In terrain navigation, traditional aircraft navigation sensors are integrated with a digital terrain map. A promising point-mass algorithm has been applied to the problem in previous work. The algorithm recursively updates the probability of the aircraft position in a finite set of positions in a mesh placed over the map area. It has proven efficient in extensive simulations when the grid mesh is chosen sufficiently dense.In order to improve the computational efficiency of the algorithm a novel approach to adaptively choose the grid point positions is presented in this work. Results from wavelet theory and numerical solution of partial differential equations are utilized to adapt the grid mesh on several dyadic scales. Initial tests indicate that the new wavelet filter achieves significant computational performance gains compared to the uniform point-mass version. This fact also makes it applicable to estimation in higer dimension.