Optimal Consensus Set for Digital Line Fitting

Rita Zrour (1) , Yukiko Kenmochi (1) , Hugues Talbot (1) , Lilian Buzer (1) , Yskandar Hamam (1, 2) , Ikuko Schimizu (3) , Akihiro Sugimoto (4)

1 LIGM - Laboratoire d'Informatique Gaspard-Monge
2 F’SATI - French South African Institute of Technology
3 TUAT - Tokyo University of Agriculture and Technology
4 NII - National Institute of Informatics

This paper presents a new method for fitting a digital line to a given set of points in a 2D image in the presence of noise by maximizing the number of inliers, namely the consensus set. By using a digital line model instead of a continuous one, we show that we can generate all possible consensus sets for digital line fitting. We present a deterministic algorithm that efficiently searches the optimal solution with the time complexity O(N2 logN) and the space complexity O(N) where N is the number of points.

Mots clés: digital geometry; optimization; consensus set; outliers

Domaines: Mathématique discrète [cs.DM]; Vision par ordinateur et reconnaissance de formes [cs.CV]

Rita Zrour, Yukiko Kenmochi, Hugues Talbot, Lilian Buzer, Yskandar Hamam, et al.. Optimal Consensus Set for Digital Line Fitting. 13th International Workshop on Combinatorial Image Analysis (IWCIA'09), Nov 2009, Cancun, Mexico. pp.93-105. ⟨hal-00437456⟩