IDAPS Color Film Characterization: Draft for Review
Abstract
The focus of this effort was to determine the utility of color imagery in the GADS environment. In particular, we were looking for improved edge extraction. The underlying assumption driving this investigation was that desirable edges could be detected using color when their presence was either weak or imperceptible in the monochrome imagery. Presumably, the stronger edges from color exploitation would enhance the performance of Auto-GADS, or at least aid an operator in manual operations. It is reasonable to assume that along with more desirable edges there will be an increased presence of undesirable edges, i.e., clutter. The sensitivity of GADS operations to clutter were not addressed in this particular effort. GADS operations rely upon matching a wire frame model to edges in the data. The accuracy requirements of this match are quite strict. Thus, any bias or distortion of extracted edges arising from the color exploitation process could result in a failure to meet accuracy specifications. While this possibility appeared remote, a significant portion of this effort was directed towards determining if any shifts were present and the nature of their origin.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 1991
- Accession Number
- ADA242681
Entities
People
- Charles Lake
- David Symanow
- John Hemdal
Organizations
- Environmental Research Institute of Michigan