Accurate Junction Detection and Characterization in Natural Images

Gui-Song Xia (1) , Julie Delon (2) , Yann Gousseau (3)

1 State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing
2 MAP5 - UMR 8145 - Mathématiques Appliquées Paris 5
3 LTCI - Laboratoire Traitement et Communication de l'Information

Accurate junction detection and characterization are of primary importance for several aspects of scene analysis, including depth recovery and motion analysis. In this work, we introduce a generic junction analysis scheme. The first asset of the proposed procedure is an automatic criterion for the detection of junctions, permitting to deal with textured parts in which no detection is expected. Second, the method yields a characterization of L-, Y- and X- junctions, including a precise computation of their type, localization and scale. Contrary to classical approaches, scale characterization does not rely on the linear scale-space, therefore enabling geometric accuracy. First, an {\it a contrario} approach is used to compute the meaningfulness of a junction. This approach relies on a statistical modeling of suitably normalized gray level gradients. Then, exclusion principles between junctions permit their precise characterization. We give implementation details for this procedure and evaluate its efficiency through various experiments.

Mots clés: a-contrario method; contrast invariance; Junction detection; scale characterization

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]

Gui-Song Xia, Julie Delon, Yann Gousseau. Accurate Junction Detection and Characterization in Natural Images. International Journal of Computer Vision, 2014, 106 (1), pp.31-56. ⟨10.1007/s11263-013-0640-1⟩. ⟨hal-00631609⟩