Title :
The Euler feature vector
Author :
Stavrianopoulou, A. ; Anastassopoulos, Vassilis
Author_Institution :
Electron. Lab., Patras Univ., Greece
Abstract :
Three topological quantities are proposed in order to be used in combination with the Euler number in shape classification procedures. These quantities are evaluated using the convex deficiency of the shape as well as orientation dependent variations of its convex hull. The proposed topological feature vector is prone to problems or instabilities imposed by the rectangular grid. The computational load is related to the time needed for convex hull and Euler number evaluation. Its classification results are satisfactory when it is used in separating typewritten letters or a small number of binary silhouettes
Keywords :
image classification; topology; Euler feature vector; Euler number; binary silhouettes; convex deficiency; convex hull; orientation dependent variations; shape classification procedures; topological feature vector; typewritten letters; Airplanes; Gravity; Laboratories; MATLAB; Physics; Shape; Testing;
Conference_Titel :
Pattern Recognition, 2000. Proceedings. 15th International Conference on
Conference_Location :
Barcelona
Print_ISBN :
0-7695-0750-6
DOI :
10.1109/ICPR.2000.903719