DocumentCode
442771
Title
The Euclidean k-distance transformation in arbitrary dimensions: a separable implementation
Author
Cuisenaire, Olivier
Author_Institution
Signal Process. Inst., Ecole Polytech. Fed. de Lausanne, Switzerland
Volume
2
fYear
2005
fDate
11-14 Sept. 2005
Abstract
The signed k-distance transformation (k-DT) computes the k nearest prototypes from each location on a discrete regular grid within a given D dimensional volume. We propose a new k-DT algorithm that divides the problem into D 1-dimensional problems and compare its accuracy and computational complexity to the existing raster-scanning and propagation approaches.
Keywords
computational complexity; image processing; Euclidean k-distance transformation; raster-scanning approach; Computational complexity; Grid computing; Iterative algorithms; Military computing; Prototypes; Signal processing algorithms; Sorting;
fLanguage
English
Publisher
ieee
Conference_Titel
Image Processing, 2005. ICIP 2005. IEEE International Conference on
Print_ISBN
0-7803-9134-9
Type
conf
DOI
10.1109/ICIP.2005.1530141
Filename
1530141
Link To Document