Technologies and Algorithms for Building the Augmented Reality

Abstract

This article is about the Augmented Reality technology itself and its current implementations. In the first part of the article the authors give a short historical reference to the origins of the name ”augmented reality”, by whom it was introduced and what it means. Later in the article two major approaches to building AR are described. The first one is based on the usage of a marker, and the second one is marker-free. The first approach is examined in detail. In order to analyze video stream and recognize known objects in it, algorithms of the Computer Vision are used. The authors give a short description and the main characteristics only of two of them: genetic algorithms and feature detection & description. For a programmatic implementation of those algorithms one can use special libraries like OpenCV and AForge.NET, also mentioned in the article. Both of them give vast functional capabilities in image processing and object recognition. At the end of the article is given an example of creating AR using the OpenCV library. Main attention is payed to the problem of making projection of a 3D model on the marker’s plane. This example can be used as the foundation for a custom AR framework.

Keywords

• дополненная реальность
• компьютерное зрение
• feature detection
• маркеры
• OpenCV