In this paper the performance of passive range measurement imaging using stereo technique in real time applications is
described. Stereo vision uses multiple images to get depth resolution in a similar way as Synthetic Aperture Radar (SAR)
uses multiple measurements to obtain better spatial resolution.
This technique has been used in photogrammetry for a long time but it will be shown that it is now possible to do the
calculations, with carefully designed image processing algorithms, in e.g. a PC in real time.
In order to get high resolution and quantitative data in the stereo estimation a mathematical camera model is used. The
parameters to the camera model are settled in a calibration rig or in the case of a moving camera the scene itself can be
used for calibration of most of the parameters. After calibration an ordinary TV camera has an angular resolution like a
theodolite, but to a much lower price.
The paper will present results from high resolution 3D imagery from air to ground. The 3D-results from stereo
calculation of image pairs are stitched together into a large database to form a 3D-model of the area covered.