A multi-baseline elevation reconstruction algorithm based on closed-form clustering and unscented information filter

Aiming at the difficulty of low terrain reconstruction efficiency of interferometric synthetic aperture radar in complex terrain, a multi-baseline elevation reconstruction algorithm based on closed-form clustering and unscented information filter was presented. Firstly, this algorithm roughly estimated the terrain height using a closed-form clustering algorithm, and then obtained the terrain abrupt areas using the phase residue recognition, edge detection operator and other auxiliary information. Secondly, an unscented information filtering phase unwrapping (UIFPU) model with better anti-noise performance was constructed, and an efficient heap-sort path-following strategy was used to guide the path of phase unwrapping of the UIFPU model. Finally, the final estimated elevation of each pixel was determined according to comparing probability density functions of the estimated phase by using different approaches. The experimental results show that the proposed algorithm is efficient and can deal with the problem of terrain rapid elevation reconstruction in the case of large-scale interference data sets, and the feasibility of the proposed algorithm is fully verified by experimental results.