Closed-loop SRME Based on the L1 Norm Sparse Constraint

We proposed an improved method based on Closed-loop SRME method, modify traditional Closed-loop SRME and use sparse transform to sparse inversion of biconvex optimization, which increase the accuracy of inversion result and deep event are more continuous than before. Primary and surface operator could be estimated directly by the method we proposed iteratively. It also avoids multiple subtraction and damage to primary energy by SRME.Synthetic data will show the comparisons between Closed-loop SRME based on the L1 norm sparse constraint and other method of multiple attenuation.