dengchengcheng / ocv_ar

OpenCV based Augmented Reality library

ocv_ar - OpenCV based Augmented Reality library

Markus Konrad [email protected], June 2014

INKA Research Group / Project MINERVA, HTW Berlin - http://inka.htw-berlin.de/inka/projekte/minerva/

This project implements a simple but powerful Augmented Reality solution based on OpenCV. It is based on the explainations from the book Mastering OpenCV with Practical Computer Vision Projects (Baggio et al., 2012) and borrows some ideas and code from ArUco. It is written in C++ and has no dependencies besides the already mentioned OpenCV.

Features

• fast, lightweight and portable library in standard C++
• depends only on OpenCV library
• different marker identification possibilities:
• ArUco-style 7x7 code markers
• custom (binary) markers via template matching
• marker tracking with marker pose interpolation for smooth marker motions (see this example project)
• well documented code and example projects (see below)

Usage and example projects

The library usage is best explained with the well documented example projects, which are available for different plattforms in a separate github repository 'ocv_ar-examples'.

The library is also part of ocv_ar-cocos2d, a mobile augmented reality framework for iOS that uses Cocos2D for 3D visualization and user interaction.

Acknoledgements and Contributions

Thanks to Alexander Godoba for the 3D pose smoothing functions (especially the 3D rotation interpolation functions).