Improvements to the Gray-Level Co-occurrence Matrix (GLCM) Based Trackability Metric
Abstract
The Gray-Level Co-occurrence Matrix (GLCM) based Trackability Metric has been used at AMCOM for many years for autotracker performance evaluation. The origins of the metric stem from the frustrating experience of trying to use such measures as delta T and signal to noise ratios to specify autotracker performance requirements. This paper presents the most recent developments and improvements to the GLCM based Trackability Metric (TM) and the resulting performance enhancements. One of the new developments is the inclusion of a hot spot or most predominant feature metric. Additionally, the results of a sensitivity study for the accuracy of target segmentation verses the GLCM TM performance is presented. These new developments for the Trackability Metric should provide better state-of-the-art performance prediction and more accurate performance modeling for specific imaging autotracker designs. The results of these studies and the final implementation of the Trackability Metric are presented.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 2002
- Accession Number
- ADP023093
Entities
People
- Mark A. Strauser
- Ricky K. Hammon