An Application of Statistics to Algorithm Development in Image Analysis
Abstract
We gathered and processed infrared (IR) images taken by a quantum well infrared photodetector (QWIP) camera to support the development of an algorithm for airborne target detection. This report addresses the statistical aspects of this effort. Digitized images consisted of background clutter alone and background clutter with a small aircraft target. We used a template-matching technique to detect the small aircraft amidst the background clutter. We propose a reasonable statistic for use in our algorithm. Our primary focus is to describe the experimental design and analysis used in determining parameter settings for the detection algorithm and to report those results.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 2001
- Accession Number
- ADA395055
Entities
People
- Barry Bodt
- David Hillis
- Philip David
Organizations
- United States Army Research Laboratory