Miroshnichenko N. Computer models and methods for three-dimensional visualization of photogrammetric data

The thesis is devoted to the development of computer models and computational methods for three-dimensional visualization of photogrammetric data. The analysis of modern means of obtaining and processing of photogrammetric data for three-dimensional visualization of terrain models. Dedicated general stages of information processing and principles of spatial models using computer graphics. A new classification of objects in geometric modeling systems. A new mathematical model of three-dimensional visualization of spatial data, based on the gradual application of the method of pre-processing of digital data based on Wiener filter and the method of adding profiles gamut images, thus improving the quality of visualization of three-dimensional surface interpolation stage. The method of managing spatial data visualization based on Bicubic spline method to change the structure of the model without changing the structure of its control points. A shader optimization cost computing resources based on the proposed method of treatment metatrees digital spatial data that allow us to visualize models containing more than 2 million pixels. Based on the proposed algorithms and models developed understanding of three-dimensional objects created software system of interactive processing and visualization. The basis of these methods use the full cycle laid graphics pipeline. Practical results form the prerequisites for solving applied problems of construction and visualization of three-dimensional spatial models of digital data with a lack of a priori information. The results of the research implemented in practice and in the learning process.