Title :
Spiral symmetry centroid detection
Author :
Jiandong Wang ; Wei Feng ; Jizhou Sun
Author_Institution :
Sch. of Comput. Sci. & Technol., Tianjin Univ., Tianjin, China
Abstract :
Radial symmetry detection is a hot topic in computer vision owing to its importance in understanding the image structure and content. A major generalisation to the problem is presented and single image spiral symmetric centroid detection is studied. Since radial symmetry is just a particular case in the model of spiral symmetry, spiral symmetry detection is more challenging and general. For this new problem, a fast and effective voting-based solution is proposed, which can reliably detect both the classical radial symmetric centre and the more general spiral symmetric centroid. Extensive experiments validate the generality and practicability of the proposed approach.
Keywords :
computer vision; object detection; computer vision; image structure; radial symmetry detection; single image spiral symmetric centroid detection; voting-based solution;
Journal_Title :
Electronics Letters
DOI :
10.1049/el.2014.0601