稀疏网格正交非线性滤波多基线相位解缠算法OA

The multi-baseline phase unwrapping method can effectively solve the elevation inversion problem of discontinuous or abrupt terrains,and a multi-baseline phase unwrapping algorithm based on deep learning for sparse grid quadrature nonlinear filters was proposed to solve the problem of weak robustness of the multi-baseline elevation reconstruction method.First,a multi-channel edge detection network was built to extract edge information from multi-channel interferograms.Second,the interferograms were divided into mutation regions and continuous regions according to the edge information,and the elevation was acquired by the total variation maximum a posterior algorithm for the mutation regions,and the sparse-grid quadrature nonlinear filter was used to predict the continuous regions.Experimental results show that the proposed method achieves good elevation reconstruction accuracy and noise robustness.