This module produces an output intended for machine use, and extracts cropped images around the N most salient locations. N is decided by the height of the desired USB output image. See DemoSaliency for more information about visual attention and saliency.
Here we show how the output image size chosen in the videomappings.cfg config file can be used to decide what output to compute. Namely, here we assume square regions stacked vertically. So the width of the outut image sets region size, and the height divided by that size sets the number of regions.