作者: Freek Reinders , Melvin E. D. Jacobson , Frits H. Post
DOI: 10.1007/978-3-7091-6783-0_8
关键词: Feature (computer vision) 、 Feature extraction 、 Graph (abstract data type) 、 Pattern recognition 、 Topological graph 、 Topological skeleton 、 Spatial network 、 Mathematics 、 Orientation (graph theory) 、 Artificial intelligence 、 Morphological skeleton
摘要: An essential step in feature extraction is the calculation of attribute sets describing characteristics a feature. Often, include position, size, and orientation These attributes are very important, but they do not provide good approximation shape For better description, more sophisticated method needed. This paper describes that extracts binary skeleton feature, transforms it into graphical representation: skeletongraph. graph represents original with controlled precision, contains topology geometry skeleton. In addition, distance information used to generate simplified reconstruction 3D shape, which can also be as an iconic object for visualization.