Conclusions

To avoid distortion in the image, 3-D ADCIGs must be properly weighted before averaging over azimuths. These weights can be derived from the jacobian of the transformation to angle domain, after taking into account the folding of the azimuth axis. The weighting method I presented was successful to correct the image obtained from a synthetic data set.

To attenuate both coherent noise and numerical artifacts, I also defined a simple scheme for limiting the azimuthal range of the image. This variable azimuthal range proved to be useful in reducing the artifacts caused by the limited azimuthal coverage of the synthetic data. Further analysis of the artifacts caused by the azimuthal coverage of real acquisition geometries might be required to make the computation of 3-D ADCIGs a robust tool for velocity analysis and amplitude analysis.