Surface Wrapping
Surface Wrapping is a tool I developed for my doctoral thesis that can be used to capture complex three-dimensional surfaces from noisy volumetric image data. It is a semi-automated technique that relies on a unique interface that allows the user to efficiently define an approximate bounding surface in 3D, creating a mesh that is then algorithmically deformed under the user's control to create the best fit with the data.

Surface Wrapping has evolved from an academic research topic to a successful commercial product, sold as a part of the InsightEarth suite from CGG. It was primarily developed for geologic interpretation of seismic data, but also has applications in medical image segmentation, fitting surfaces to point cloud data, and more.

Download my dissertation (PDF, 29.5MB)

Inspiration
Surface Wrapping was derived from the Surface Draping horizon interpretation method created by Dr. Geoffrey Dorn, and is actually a more flexible generalization of that technique. The concept is inspired by shrink wrapping: a deformable sheet or bag is loosely wrapped around an object, then shrunk until the desired fit is achieved.


Depth-shaded rendering of a complex dewatered shale surface.
Examples of results
The original goal of Surface Wrapping was to allow rapid interpretation of channels, but it has proven to be useful for many other purposes. In addition to picking objects in seismic data such as salt bodies, horizons, and canyon and reef boundaries, it has also been applied to 3D medical data from CT and MR scans. The approach is very tolerant of data that are noisy or poorly-imaged, making it possible to achieve accurate results quickly when working with low-quality sources.