A Two-Stage Classification Approach for Effective Geomorphic Mapping of Planetary Surfaces

Tomasz F. Stepinski1, Chaitanya Bagaria2
1 Lunar and Planetary Institute, Houston, TX 77058, USA
Telephone: (+1-281-486-2170)
Fax: (+1-281-4862162)
Email: tom@lpi.usra.edu
2 Dept. of Computer Science, University of Houston, 4800 Calhoun Rd., Houston, TX 77204
Email: chaitanya.bagaria@gmail.com

Advances in remote sensing from spacecrafts have produced a large amount of data on topography of planetary surfaces. In particular, the entire surface of planet Mars is covered by a digital elevation model (DEM) with a resolution of ~500 m derived from laser altimeter measurements. In addition, an increasing number of sites on Mars are covered by higher resolution DEMs derived from stereo images. The high resolution global DEMs of planet Mercury and the Moon will be available in the near future. Last but not least, most landmasses on Earth are covered by the 30-90 m/pixel DEM produced from data collected by the Shuttle Radar Topography Mission (SRTM). The major tools for understanding the origin and evolution of planetary surfaces are geomorphic and geologic maps that are traditionally created manually on the basis of photo-geologic interpretation. The slowness and expense of manual methods severely limits the area that can be mapped at the level of detail corresponding to the resolution of available elevation data. For example, 1:500,000 geomorphic maps of Mars exist only for a tiny percentage of its surface. Thus, there is a critical need to develop an effective method for automating the process of geomorphic mapping. In this paper we describe a framework for auto-generation of such maps. The resultant maps have information esthetics similar to manually drawn maps and they can be stored in a standard GIS shapefile format. We assert that our method has a combination of features that makes it likely to become a useful exploratory tool for planetary scientists.

AttachmentSizeHitsLast download
stepinski2009geomorphometry.pdf1.51 MB198720 hours 19 min ago
4.4
Your rating: None Average: 4.4 (5 votes)