The Euclidean distance transform in arbitrary dimensions

Abstract The original sequential Euclidean distance transformation is not separable. This makes it useful only on single processor systems. We suggest variants for 2, 3 and arbitrary dimensions that are separable, suitable for various parallel architectures. The results include a 4-scan algorithm for 3-dimensional images.

[1]  G. Borgefors.  Distance transformations in arbitrary dimensions , 1984 .

[2]  Azriel Rosenfeld,et al.  Sequential Operations in Digital Picture Processing , 1966, JACM.

[3]  Gunilla Borgefors,et al.  Distance transformations in digital images , 1986, Comput. Vis. Graph. Image Process..

[4]  P. Danielsson.  Euclidean distance mapping , 1980 .

[5]  Ugo Montanari,et al.  A Method for Obtaining Skeletons Using a Quasi-Euclidean Distance , 1968, J. ACM.

[6]  Ruzena Bajcsy,et al.  Packing Volumes by Spheres , 1983, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[7]  Q. Ye.  The signed Euclidean distance transform and its applications , 1988, [1988 Proceedings] 9th International Conference on Pattern Recognition.