| Autor: |
BURDESCU, Dumitru Dan, EBANCA, Daniel Costin, SLABU, Florin |
| Předmět: |
|
| Zdroj: |
International Journal of Computer Science & Applications; Aug2015, Vol. 12 Issue 2, p120-143, 24p |
| Abstrakt: |
The goal of this paper is to present two methods of segmentation on color images based on graphs algorithms. The paper contains two parts. The first part presents Graph-Based Segmentation Method for Planar Images and despite of the majority of the segmentation methods our method does not require any parameter to be chosen in order to produce a better segmentation. A virtual hexagonal structure is created on the pixels of the initial image, and also the initial triangular grid graph having hexagons as vertices. Then the color-based sequence of segmentations and its associated sequence of forests are generated by using the color-based region model and a maximum spanning tree construction method based on a modified form of the Kruskal's algorithm. The syntactic-based sequence of segmentations and its associated sequence of forests are generated by using the syntactic-based model and a minimum spanning tree construction method based on a modified form of the Boruvka's algorithm. Finally, color and geometric features are extracted from the determined regions. The second part presents Graph-Based Segmentation Method for Spatial Images. We shall present a unified framework for spatial digital image segmentation and contour extraction that uses a virtual tree-hexagonal structure (prism cells) defined on the set of the input digital image voxels. The advantage of using a virtual tree-hexagonal network superposed over the initial image voxels is that it reduces the execution time because we are using graphs algorithms and their complexities. So, our algorithms are linear on the input the number of the vertices of the spatial input graph. The key to the whole algorithms of spatial segmentation method is the prism cells. This segmentation method contains many other algorithms but only graph-based segmentation algorithm is presented based on the limited space of paper. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
