Measurements and Modeling Enhancements for the NPS Minimum Resolvable Temperature Difference Model, VISMODII
Abstract
Minimum Resolvable Temperature Difference (MRTD) measurement has long been used to describe the performance of thermal imaging systems Computer models such as U,S Army's FLIR92, that were developed to predict the MRTD, were reported to have deficiencies in dealing with sampling and aliasing effects, The models also include assumptions regarding the observer recognition process and therefore cannot predict the MRTD of an imager that incorporates an "objective" automatic target recognition device instead of a "subjective" human observer, The Visibility Model II developed for second generation thermal imaging systems at the Naval Postgraduate School (NPS) in the mid 90's takes sampling and aliasing issues into account and makes no assumptions about the observer, Modeling enhancements in VISMODII and its extension to predict objective MRTD are proposed and tested in this thesis, A parallel thesis at the NPS has shown that aliasing effects on image appearance are fundamentally different from noise, The improved VISMODII model accounts for the fact that unlike noise, aliasing may have a visual enhancing effect and therefore may lower MRTD, Experiments were conducted to measure subjective and objective MRTD, Experimental results demonstrated that the VISMODII model successfully predicts the MRTD both for the subjective and the objective schemes
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 2001
- Accession Number
- ADA397426
Entities
People
- M. Celik
Organizations
- Naval Postgraduate School