Authors:
Nicolas Coudray
1
;
Argyro Karathanou
1
and
Sylvie Chambon
2
Affiliations:
1
Université de Haute-Alsace, France
;
2
Laboratoire Central des Ponts et Chaussées, LCPC, France
Keyword(s):
Segmentation, Multi-resolution, Road images, Cracks.
Related
Ontology
Subjects/Areas/Topics:
Computer Vision, Visualization and Computer Graphics
;
Feature Extraction
;
Features Extraction
;
Image and Video Analysis
;
Informatics in Control, Automation and Robotics
;
Segmentation and Grouping
;
Signal Processing, Sensors, Systems Modeling and Control
Abstract:
In the context of fine structure extraction, this paper presents a new method based on multi-resolution segmentation applied for the detection of road cracks. A method already developed to detected low-contrasted biological membranes has been adapted to detect cracks on images: crack features are defined as heterogeneities rather than transitions of closed regions characterizing the membranes. This new methodology is quantitatively validated on reference segmentations and compared to an adapted filtering and Markovian modelling algorithm.