Sequences of Moving Targets

The experimental motion sequences were created from 44 slides taken during a field test exercise (the Distributed Interactive Simulation, Search and Target Acquisition Fidelity field test), that was designed and organized by NVESD (Night Vision & Electro-optic Sensors Directorate, Ft. Belvoir, VA, USA) and that was held in May and June 1995 in Fort Hunter Liggett, California, USA $\cite{toet:1}$.

These slides depict 44 different scenes (see [10]). Each scene represents a military vehicle in a complex rural background. The 9 different vehicles that are deployed as search targets are respectively a BMP-1, a BTR-70, an HMMVV-Scout, a HMMVV-Tow, an M1A1, an M3-Bradley, an M60 , an M113, and a T72. The visibility of the targets varies throughout the entire stimulus set.

For each image, containing a target (vehicle), a corresponding background scene was created. The background scene should be everywhere equal to the target scene, except at the location of the target, where the target support is filled with the local background. This replacement was done by hand, using the rubber stamp tool in a commercial software.

The intent of this operation is to create an image identical to the target scene outside the target region, and in which the contrast gradients for the target region are induced by the background bounding the target. The background image is thus obtained by smoothing the values in the background at the target surrounding area. This smoothing produces the blending of the local surround into the target region. The result is judged by eye and is accepted if the variation in the background over the target support area does not appear to have an appreciable contrast with the natural variation in the local background.

The original sequences of 16 frames (with a frame size of $256 \times 256$) are then developed using a proper tool to place the targets into the background scenes. The target starting and ending locations are calculated in order to represent the appropriate velocities (three different speeds) for each specified range (the sequences represent different ranges). The targets started at either a left, central or right grid for a given range. The direction is determined randomly.

In the experimental sequences of moving targets there are also secondary environmental effects (e.g., motion-of-vegetation) caused by simulated camera motion, with the intent that they correspond to the true situation of near-identical frames in a field situation.