Enhanced omnidirectional image unwrapping for face detection

This paper introduces a new framework to improve the performance of Viola and Jones face detector on omnidirectional unwrapped images. First, an optimization scheme is used to improve the unwrapped image specifically for rectangular Haar-like features. Then, we compare our unwrapping approach to the performance obtained with spherical unwrapping. The impact of the decision boundary and candidate window density are also investigated. Our work suggests that our new unwrapping technique improves significantly the performance of Viola and Jones detector on omnidirectional unwrapped images.

[1]  Yohan Dupuis,et al.  A framework for face detection on Central Catadioptric Systems , 2010, 2010 IEEE International Workshop on Robotic and Sensors Environments.

[2]  Pascal Vasseur,et al.  Motion estimation by decoupling rotation and translation in catadioptric vision , 2010, Comput. Vis. Image Underst..

[3]  Eric Sommerlade,et al.  Hidden view synthesis using real-time visual SLAM for simplifying video surveillance analysis , 2011, 2011 IEEE International Conference on Robotics and Automation.

[4]  Andre L. C. Barczak,et al.  Face tracking using a hyperbolic catadioptric omnidirectional system , 2009 .

[5]  Yohan Dupuis,et al.  Robust Radial Face Detection for Omnidirectional Vision , 2013, IEEE Transactions on Image Processing.

[6]  Andreas Zell,et al.  Robust real-time detection of multiple balls on a mobile robot , 2013, 2013 European Conference on Mobile Robots.

[7]  Simone Gasparini,et al.  Camera Models and Fundamental Concepts Used in Geometric Computer Vision , 2011, Found. Trends Comput. Graph. Vis..

[8]  Paul A. Viola,et al.  Fast Multi-view Face Detection , 2003 .

[9]  Maxime Lhuillier.  Toward automatic 3D modeling of scenes using a generic camera model , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.