A thinning algorithm is used to reduce unnecessary information by peeling objects layer by layer so that the result is sufficient to allow topological analysis. It has several applications, but is particularly useful for skeletonization. A new sequential thinning algorithm is introduced to preserve both the topology and geometry of the object. In sequential thinning, only a single point may be deleted at a time and it always guarantees the preservation of the topology of the original image. It is based on removing the central pixel in the 3x3 neighborhood of the candidate pixel which preserves the topology and geometry. The algorithm is based on computing the local Euler number before and after removing the candidate pixel and then checking whether there is a difference in the computed values. Furthermore, in order to preserve the geometric criteria (preserve end points of the object), we consider change of the boundary length when removing the central pixel.
[1]
Joachim Ohser,et al.
THE EULER NUMBER OF DISCRETISED SETS – SURPRISING RESULTS IN THREE DIMENSIONS
,
2011
.
[2]
Kristel Michielsen,et al.
Morphological image analysis
,
2000
.
[3]
Jean Serra,et al.
Image Analysis and Mathematical Morphology
,
1983
.
[4]
Pierre Soille,et al.
Morphological Image Analysis: Principles and Applications
,
2003
.
[5]
Hamid Krim,et al.
Object Recognition Through Topo-Geometric Shape Models Using Error-Tolerant Subgraph Isomorphisms
,
2010,
IEEE Transactions on Image Processing.
[6]
Gabriella Sanniti di Baja,et al.
Computing skeletons in three dimensions
,
1999,
Pattern Recognit..
[7]
Joachim Ohser,et al.
The Euler Number of Discretized Sets — On the Choice of Adjacency in Homogeneous Lattices
,
2002
.