Reuta O. Geometric models for analysis and interpretation of projective image shapes

The grey scaled images of plane objects obtained as central projections in conditions of a distant photogrammetry. The in-crease of efficiency of the image identification and automated analysis that is based on a multidimensional model for the geomet-rical shapes of central projections meaning inhomogeneous distri-bution of its brightness. Methodical, algorithmic and program maintenances of positional-invariant image recognition on the basis of offered informational-geometrical models which is pre-sented in terms of position and metric problems of multidimen-sional geometry of its information characteristic space are de-signed. The algorithms of identification and image analysis are proposed. Its high performance is justified theoretically. They have been applied in such directions: the automated analysis of the grey scaled photogrammetric images of surface regions obtained from satellite launchers; calculations of quantitative measures of difference of two images for one fixed object obtained under di f-ferent conditions of projection; recovery of projection parameters