Automated Digital Elevation Extraction from Multispectral Imaging Systems

NASA is seeking improved capabilities to help make Earth measurement easier to use and more reliable for practitioners. Accordingly, we propose to research advanced methods for generating automatically precise elevation data from modern high spatial resolution digital stereoscopic panchromatic/multispectral sensors using improved digital auto-correlation techniques. Applications for this solution include medium-to-large-scale mapping (e.g., community development and resource/environmental management), planning (e.g., event modeling, urban, and mission), and assessment (e.g., post-disaster reporting). Highly accurate digital elevation matrices are required to generate reference data having:- Results needing little interactive editing;- Reliable and valid elevation information;- Postings ranging from 3m-to-cm; and- Abilities to maximize the utilities provided by modern sensors.Automatic digital auto-correlation for elevation data is critical for a variety of users. Current solutions, however, have been limited due to a general lack of algorithms addressing the terrain complexities brought forth in high-spatial resolution data. (They are also typically difficult to use.) This includes generation of reference data for elevation profiles supporting vertical update point technologies and related terrain and feature models. Furthermore, the commercial and defense mapping marketplaces, computer vision industries, and the global medical imaging communities seek similar modeling solutions, which bodes well for eventual commercialization.