Rendering obrazu rasterizací na grafické kartě

Abstract

The goal of this diploma thesis is to create a computer program that uses input data describing a graphic scene to render an image to a file. As the rendering method rasterization has been chosen. The method is processed by two distinct hardware components - processor and graphics card. The program is written in C++ programming language, mostly using its standard libraries. For processing on the graphics card NVIDIA CUDA technology is used to achieve the goal of maximum possible parallelization and therefore significantly increased speed of performing the task. The conclusion of the diploma thesis shows the comparison between both standard and parallel approach.