We have presented a methodology for performing image-domain wavefield tomography.
Our method is extremely efficient, because we formulate the velocity estimation problem in a target-oriented fashion,
and use a synthesized Born data set for velocity inversion.
Our method is wavefield-based; therefore it is suitable
for estimating velocities in complex geologies.
Numerical examples on a 3-D field data set
give excellent inversion results.
The entire velocity analysis workflow can be fully automated, and no picking is necessary,
although human interaction can be easily incorporated if it is desired.
Thus our method provides a good tool for fast, accurate and flexible velocity model building in areas
with complex geologies.