Quick Search:    advanced search

GSW Home   GeoRef Home   My GSW Alerts   Contact GSW   About GSW   Journals List   Help

This Article Figures Only Full Text Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Levin, S. A. Articles by de Hoop, M. Search for Related Content GeoRef GeoRef Citation

Extracting information from geophysical, medical, and space images

Landmark Graphics, Denver, Colorado, U.S.

Martijn de Hoop

Colorado School of Mines, Golden, Colorado, U.S.

Corresponding author: SALevin@lgc.com

The first 20% of the full text of this article appears below.

The fourth session of the 2001 SEG Summer Research Workshop, "Image Processing II—Extracting Information from Images," complemented "Image Processing I—Forming and Enhancing Images" by focusing on methods for analyzing images for important features and extracting qualitative and quantitative results. All three disciplines that participated in the workshop—geophysics, medicine, and space-based remote sensing—have an escalating need to automatically extract information from digital images and speakers in all three fields testified to the commonality of techniques and synergies among disciplines.

	Improving images by analyzing boundaries

 
Several presentations discussed how to find boundaries between distinct components of an image—a field known as image segmentation. Leadoff speaker Ravi Malladi (Lawrence Berkeley National Laboratory) showed striking results in applying fast, elegant PDE-based level-set methods in medical, biomedical, and geophysical settings to shape segmentation, image noise removal, and tracking of moving targets. Malladi drew his examples from both medical and geophysical imagery, including 3D feature tracking. These methods were first developed by James Sethian at Berkeley and Stanley Osher at UCLA to study phenomena that have moving boundaries, such as combustion, and these methods have been adapted within the geophysical community to accomplish fast and accurate calculation of seismic traveltimes. In this presentation, applications included space research and geophysics. The methodology in Malladi's talk is a potentially powerful alternative to the watershed algorithm (a method in which paths are conceptually followed "downhill" from each image pixel, assigning two points to the same image component if their corresponding paths meet) and we look forward to seeing synergies develop in this area of image segmentation.

In some examples from Malladi's talk, Figure 1 shows noise reduction and texture enhancement of volumetric medical images using Beltrami flow. This is one of a class of algorithms in which a partial differential equation represents an expanding front. The equation is constructed so as to slow the . . . [Full Text of this Article]