Mosaicing

Video is big in terms of file size, and while video is very good at identifying motion, it is very bad at providing good overall situational awareness. Many times it is desirable to create large still images from the sequence of video frames contained in a motion imagery clip. We call that process mosaicing.

By using highly accurate image registration, 2d3 can stitch together adjacent frames, and use them to create a large mosaic - much like a video paint brush. Still mosaics have many uses, not the least of which is that a still image is much easier to transmit over a limited bandwidth datalink, than the video that produced it. This allows for an analyst to utilize motion imagery without having to look at each and every frame.

Mosaicing also allows for other capabilities, such as object segmentation to be brought to bear, and give an analyst the ability to create still image files with certain objects highlighted or removed. When combined with 2d3's terrain reconstruction capability, mosaicing allows an orthro-rectified texture to be applied to a 3d model to create a representation of a art of the earth.

Mosaicing can be accomplished as a post-process or in realtime.