Extracting ridges of topographic surfaces using scanline algorithm

The ridge/valley-like structures of a topographic surface have been extensively studied in image processing to obtain useful skeleton-like shape descriptors. In this paper, we propose a novel method that rapidly constructs the graph representation of the skeletons of the curvilinear objects such as...

Full description

Bibliographic Details
Main Author: Sukmoon Chang
Format: Article
Language:English
Published: Stefan cel Mare University of Suceava 2009-01-01
Series:Journal of Applied Computer Science & Mathematics
Subjects:
Online Access:http://www.jacs.usv.ro/getpdf.php?issue=6&paperid=63
Description
Summary:The ridge/valley-like structures of a topographic surface have been extensively studied in image processing to obtain useful skeleton-like shape descriptors. In this paper, we propose a novel method that rapidly constructs the graph representation of the skeletons of the curvilinear objects such as vascular networks and fingerprints. The method first constructs the topographic surface of a given image in such a way that the ridges of the surface reflect the medial axis of the objects in the image. The ridge points are then identified using two orthogonal scanlines and connected by tracing the maximum gradient paths on the surface. We present the very promising results of the method applied to various binary and grayscale images to demonstrate its correctness and robustness.
ISSN:2066-4273
2066-3129