Shape recognition based on Kernel-edit distance
Computer Vision and Image Understanding
3D block-based medial axis transform and chessboard distance transform based on dominance
Image and Vision Computing
Fast and accurate approximation of digital shape thickness distribution in arbitrary dimension
Computer Vision and Image Understanding
Mathematics and Computers in Simulation
| Hi-index | 0.00 |
In this paper, we propose a new approach based on three-dimensional (3-D) medial axis transformation for describing geometrical shapes in three-dimensional images. For 3-D-images, the medial axis, which is composed of both curves and medial surfaces, provides a simplified and reversible representation of structures. The purpose of this new method is to classify each voxel of the three-dimensional images in four classes: boundary, branching, regular and arc points. The classification is first performed on the voxels of the medial axis. It relies on the topological properties of a local region of interest around each voxel. The size of this region of interest is chosen as a function of the local thickness of the structure. Then, the reversibility of the medial axis is used to deduce a labeling of the whole object. The proposed method is evaluated on simulated images. Finally, we present an application of the method to the identification of bone structures from 3-D very high-resolution tomographic images.