Authors:
Sinan Kalkan
1
;
Shi Yan
2
;
Florian Pilz
2
and
Norbert Krüger
3
Affiliations:
1
Bernstein Centre for Computational Neuroscience, Univ. of Göttingen, Germany
;
2
Medialogy Lab, Aalborg Univ. Copenhagen, Denmark
;
3
Cognitive Vision Group, Univ. of Southern Denmark, Denmark
Keyword(s):
Junction Detection, Junction Positioning, Junction Representation.
Related
Ontology
Subjects/Areas/Topics:
Computer Vision, Visualization and Computer Graphics
;
Early Vision and Image Representation
;
Feature Extraction
;
Features Extraction
;
Image and Video Analysis
;
Informatics in Control, Automation and Robotics
;
Signal Processing, Sensors, Systems Modeling and Control
Abstract:
Every junction detector has a set of thresholds to make decisions about the junctionness of image points.
Low-contrast junctions may pass such thresholds and may not be detected. Lowering the thresholds to find such junctions will lead to spurious junction detections at other image points. In this paper, we implement a junction-regularity measure to improve localization of junctions, and we develop a method to create semantic interpretations of arbitrary junction configurations at improved junction positions. We propose to utilize such a semantic interpretation as a feedback mechanism to filter false-positive junctions. We show results of our proposals on natural images using Harris and SUSAN operators as well as a continuous concept of intrinsic dimensionality.