Kucherenko V. Formalized geometric model of an irregular surface for hyper quantity discrete finite set of dots

The object is the processes of remote sensing multispectral raster images forming and pre-processing while images are got by means of scanner. The purpose is the development of theoretical base, algorithmic and program means of material object geometric forms identification on the base of multispectral images fixed in optical and infrared electromagnetic specter range provided invariance with respect to positional forming factors which are determining image spatial and radiometric resolution. The methods of geometric modeling, the object geometric forms identification, parameterization theory, projective and computing geometry, linear algebra, set theory, general information theory and mathematics statistics, and aslo image fusion, the data dimension reduce, de-correlation and geometric correction of such image distributions. Geometric modeling method and modes of photogrammetric images forming process and pre-processing for the purpose of material object geometric form identification are developed. In the dissertation the essential science and practical problem of object geometric forms identification on digital multispectral photogrammetric images using the fussy logics rules is solved. With this purpose the method rising the information value of such images according to geometric form recognition trustworthiness criterion is developed. The method provides invariance with respect to positional forming factors of multispectral images which are determining image spatial and radiometric resolution. The results are introducted in SDO of spacecrafts and systems State-run Enterprise "Design office "Yuzhnoe" named after М.К. Yangel" (Dnipropetrovsk); educational process of Dnipropetrovsk national university named after Oles Honchar. Scope - landscape data thematic processing.