Author:

Peter Kán
Faculty of mathematics, physics and informatics, Comenius University, Bratislava
peterkan@peterkan.com, www.peterkan.com

Supervised by:

Andrej Ferko
Faculty of mathematics, physics and informatics, Comenius University, Bratislava
ferko@sccg.sk

Abstract:

The automatic modeling of a 3D human head has challenged researchers in computer graphics for many years, because 3D head models are useful in several application areas to assist in different tasks. In this paper we present the novel system for automatic 3D head model creation from two images based on a parameterized head model with a hierarchical tree of facial features. The proposed system is divided into three parts. In the first one we use computer vision techniques like Skin-Tone based image segmentation or Haar Cascade Classifiers to detect head parameters for parameterized model update from frontal and profile head photograph. Then a 3D head model is reconstructed by updating a parameterized model with parameters detected from images and a texture is mapped. In the second part the reconstructed head model is rendered with real-time rendering techniques simulating skin illumination. In the third part the model is exported to the Collada format for further use in other applications. The system for automatic image-based 3D head modeling can be used e.g. in computer games development, movies, telecommunications, medicine, or security systems for human identification.

PDF: Download the full paper

Images:

Front (Left) and profile (Middle) source photographs and reconstructed 3D head model (Right)