The dynamic models explained through this paper solve the homologous image coordinates search and the image orientation for a CCD of a simplified MMS for 3D-surveys. A first dynamic model for such an orientation, extending the usual “direct method” (but with GPS measures only) with the “indirect method”, has been issued and tested some years ago. By a dynamic approach again, a matching model for an image sequence acquired with only one CCD, has been also recently proposed, so self-extracting points for relative orientation.
In this way, the flow chart for the dynamic analysis of these image sequences only involves, as user’s actions, the point collimation in the first image portraying it. In fact, first, the “matching algorithm” automatically extracts the homologous point in the second image (with external orientation only approximated by “cubic splines functions”). Therefore, the “orientation algorithm”, using coplanarity condition on such a points couple, allows to improve the external orientation from spline approximated values to better ones.
