The Split-Bregman Algorithm for Least Squares Migration in Imaging-Domain

Considerable attention has been attracted to the least squares migration(LSM), particularly in the data domain, which can mitigate partly the problems that the conventional migration has imported, such as migration artifacts and unbalanced amplitudes. However, the imaging domain least squares migration (ID-LSM) is apparently ignored. We present a new ID-LSM based on the split-Bregman algorithm. In imaging domain, conventional migration is the process of the non-stationary spacial convolution and reflection coefficient estimation is the process of the image restoration. A split-Bregman algorithm with TV norms and L1 norms regularization is utilized to restore final result. Visible advancement in 2D marine field data test has testified the validity of the ID-LSM based on the split-Bregman algorithm.