1
2
3
International journal of computer vision, 80 (2008) S. 375 - 405
4