Implementace metod snímání objektů ve 3D pro účely strojového vidění

Abstract

This thesis describes different methods which are used to acquire a 3D image projection of real word, with a main goal to use this image for a machine vision processing. Specifically, the thesis is about algorithms of triangulation, stereometry, that are based on a multipoint system, and a Time of Flight camera. Those cameras are specially designed to measure the time of traveling light to and from object, and later evaluate the real distance of the object in the space. Part of the thesis is dedicated to a mathematical model of the image processing used in computation algorithms of stereometry and triangulation, as well as a general theory of image processing. This is concluded with a deeper explanation of methods that are used for stereometry image processing. Practical part is created in development environment LabVIEW, that represent graphical programing or G language, and results of different methods are concluded based on the input arguments, and mathematically modeled complexity of evaluation process.