Matching and merging high-resolution and legacy seismic images

Conclusions

We have proposed an approach to matching seismic images of different resolutions. Our first step is non-stationary smoothing of the high-resolution image to match the spectral content and amplitudes of the legacy image. Next, we estimate the relative time shifts using local similarity scanning. After matching the two images, we create a blended image by least-squares inversion, which effectively combines the best features of the two images: the broader frequency bandwidth of the high-resolution image with the reflection continuity and deeper coverage of the legacy image. The final result is an interpretable blended image. We have shown example applications of the proposed method to high-resolution and legacy images from the Gulf of Mexico.