A number of factors have come together in recent years to make the reuse of visual databases a sought after goal. Primary among these has been the rising cost of database development due to ever larger realworld gaming areas and ever increasing system capacities. On the more positive side, the basic technology of the available image generation platforms have begun to share some common architectural features which make conversions from one format to another a somewhat more manageable task than in years past.

This paper will discuss the implementation issues, and successful completion, of one specific conversion effort. The requirement is in support of an advanced Weapons Tactics Trainer which utilizes a Lockheed Martin CompuScene VI for the out-the-window images and a Silicon Graphics RealityEngine2 for infrared imagery. Correlation between the visual and IR images is required to be as close as possible, preferably exact.

The points covered will include the development and implementation of the conversion algorithms and discussion of the numerous issues which result from supporting two platforms from different manufacturers. These include differences in texture map formats and utilization, polygon capacity, management of geo-centric databases, material code correlation and Z-buffer priority solutions. Also discussed will be the lessons learned that will be applied to future conversion efforts, including the use of SIF data for visual and sensor applications, and the practicality of developing true plug and play databases that will be compatible with a wide range of image generation platforms.