Assessing Camouflage Methods Using Textural Features
Abstract
Developments in the area of signature suppression make it progressively more difficult to recognize targets. In order to obtain a sufficient low degree of false alarms it is necessary to observe spatial and spectral properties. There is a genuine need to use spatial properties when analyzing the difference between a target area and a background area. This is more relevant today since modern signature suppression techniques have focused on the reduction of distinct features, like hot spots in the infrared band. The approach is to apply texture descriptors to characterize the background and also more or less camouflaged targets. In addition other descriptors are used to characterize man made objects. It is necessary to focus on features which discriminate targets from the background, and this demands a more precise description of the background and the targets than usual. The underlying assumption is that an area with more or less observable targets has different statistical properties from other areas. Statistical properties together with detected target specific features like straight lines, edges, corners or perhaps reflections from a window have to be combined with methods used in data fusion. Experiments with a computer program that estimates the statistical differences between targets and background are described. These differences are computed using a number of different distance measures.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 2000
- Accession Number
- ADP010540
Entities
People
- Klamer Schutte
- Sten Nyberg