Title :
Anchor point thinning using a skeleton based on the Euclidean distance transformation
Author :
Mekada, Yoshito ; Toriwaki, Junichiro
Author_Institution :
Graduate Sch. of Eng., Nagoya Univ., Japan
Abstract :
Thinning is one of the most frequently used methods to obtain the geometrical features of objects. It also provides the topological feature and length measurements about an object. The paper presents a three dimensional thinning method which can control the quality of result concerning appearance of spurious short branches by a parameter value. This method is constructed by integrating the anchor point thinning algorithm and the skeletonization algorithm based upon the Euclidean metric. We applied the proposed method to artificial figures and a three dimensional bronchus region extracted from a real chest X-ray CT image and confirmed that the proposed method could control the number of spike branches and shrinkage of branches.
Keywords :
computerised tomography; image thinning; lung; medical image processing; Euclidean distance transformation; anchor point thinning; branch shrinkage; bronchus; chest X-ray CT image; geometrical feature; length measurements; skeleton; skeletonization algorithm; spike branches; topological feature; tree structure; Computed tomography; Euclidean distance; Length measurement; Shape; Skeleton; X-ray imaging;
Conference_Titel :
Pattern Recognition, 2002. Proceedings. 16th International Conference on
Print_ISBN :
0-7695-1695-X
DOI :
10.1109/ICPR.2002.1048186