Share Email Print
cover

Proceedings Paper

Assessment of synthetic image fidelity
Author(s): Marilyn A. Gilmore; Ian R. Moorhead; Graham H. Watson; Mitchell G. A. Thomson; Tim Yates; Tomasz Troscianko; David J. Tolhurst; David R. Filbee
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

When using synthetic imagery it is essential that it is fit for purpose. Imagery can be rendered at different levels of quality, depending on application. For example, when using a real-time system, rendering speed is a critical parameter but, when assessing the effectiveness of a camouflage system, physical accuracy is likely to be more important. A method to quantify the accuracy of the imagery, for particular applications, is necessary. A range of different metrics based on wavelets, higher order statistics, and a human vision model, has been developed to assess the fidelity of synthetic imagery. These metrics have been used to analyze synthetic imagery rendered at different levels of fidelity and to compare the synthetic imagery with real- world imagery. Some of the metrics can be used to compare two spatially correlated images, whereas others can be used to assess particular characteristics of the image such as clutter level. The metrics, and first order statistics, have been incorporated into a tool box called FIRE (Fidelity Investigations and Reporting Environment). This paper will describe the metrics used and the results of analyses undertaken.

Paper Details

Date Published: 6 August 2002
PDF: 12 pages
Proc. SPIE 4718, Targets and Backgrounds VIII: Characterization and Representation, (6 August 2002); doi: 10.1117/12.478814
Show Author Affiliations
Marilyn A. Gilmore, Defence Science and Technology Lab. (United Kingdom)
Ian R. Moorhead, QinetiQ (United Kingdom)
Graham H. Watson, QinetiQ (United Kingdom)
Mitchell G. A. Thomson, Univ. of Derby (United Kingdom)
Tim Yates, Univ. of Derby (United Kingdom)
Tomasz Troscianko, Sussex Univ. (United Kingdom)
David J. Tolhurst, Univ. of Cambridge (United Kingdom)
David R. Filbee, INSYS Ltd. (United Kingdom)


Published in SPIE Proceedings Vol. 4718:
Targets and Backgrounds VIII: Characterization and Representation
Wendell R. Watkins; Dieter Clement; William R. Reynolds, Editor(s)

© SPIE. Terms of Use
Back to Top