On Visualization of Tree Structures from 3D Graph Images

摘要

3D images are widely used in many application fields such as pattern recognition and medical informatics, etc Tree-structured objects, such as bronchi or blood tubes, may be digitized as graph-oriented objects in 3D images since, for example, the tails of different branches of a tree may be too close to be judged as a part of a cycle. Inaccurate flow information or structure analysis may be yielded by such superfluous conjunctions. This paper proposes a revised skeletonization approach to extract tree-structured skeletons directly from original graph-oriented objects by using the thickness property of a given object Such tree-structured skeletons are easier to visualize and simpler to interpret