Skip to Main content Skip to Navigation
New interface
Reports (Research report)

Inference of Curvilinear Structure based on Learning a Ranking Function and Graph Theory

Seong-Gyun Jeong 1 Yuliya Tarabalka 2 Nicolas Nisse 3 Josiane Zerubia 1 
2 TITANE - Geometric Modeling of 3D Environments
CRISAM - Inria Sophia Antipolis - Méditerranée
3 COATI - Combinatorics, Optimization and Algorithms for Telecommunications
CRISAM - Inria Sophia Antipolis - Méditerranée , Laboratoire I3S - COMRED - COMmunications, Réseaux, systèmes Embarqués et Distribués
Abstract : To detect curvilinear structures in natural images, we propose a novel ranking learning system and an abstract curvilinear shape inference algorithm based on graph theory. We analyze the curvilinear structures as a set of small line segments. In this work, the rankings of the line segments are exploited to systematize the topological feature of the curvilinear structures. Structured Support Vector Machine is employed to learn the ranking function that predicts the correspondence of the given line segments and the latent curvilinear structures. We first extract curvilinear features using morphological profiles and steerable filtering responses. Also, we propose an orientation-aware feature descriptor and a feature grouping operator to improve the structural integrity during the learning process. To infer the curvilinear structure, we build a graph based on the output rankings of the line segments. We progressively reconstruct the curvilinear structure by looking for paths between remote vertices in the graph. Experimental results show that the proposed algorithm faithfully detects the curvilinear structures within various datasets.
Complete list of metadata

Cited literature [31 references]  Display  Hide  Download
Contributor : Nicolas Nisse Connect in order to contact the contributor
Submitted on : Thursday, October 15, 2015 - 11:35:44 AM
Last modification on : Wednesday, October 26, 2022 - 8:14:50 AM
Long-term archiving on: : Thursday, April 27, 2017 - 12:14:40 AM


Files produced by the author(s)


  • HAL Id : hal-01214932, version 1


Seong-Gyun Jeong, Yuliya Tarabalka, Nicolas Nisse, Josiane Zerubia. Inference of Curvilinear Structure based on Learning a Ranking Function and Graph Theory. [Research Report] RR-8789, Inria Sophia Antipolis. 2015. ⟨hal-01214932⟩



Record views


Files downloads