Time of Flight Image Segmentation through Co-Regularized Spectral Clustering
Time of Flight (TOF) cameras generate two simultaneous images, one for intensity and one for range. This allows tackling segmentation problems where the information pertaining to intensity or range alone is not enough to extract objects of interest from a 3D scene. In this paper, we present a spectr...
Guardado en:
| Autores principales: | , |
|---|---|
| Formato: | Objeto de conferencia |
| Lenguaje: | Inglés |
| Publicado: |
2014
|
| Materias: | |
| Acceso en línea: | http://sedici.unlp.edu.ar/handle/10915/54575 |
| Aporte de: |
| Sumario: | Time of Flight (TOF) cameras generate two simultaneous images, one for intensity and one for range. This allows tackling segmentation problems where the information pertaining to intensity or range alone is not enough to extract objects of interest from a 3D scene. In this paper, we present a spectral segmentation method that combines information from both images. By modifying the affinity matrix of each of the images based on the other, the segmentation of objects in the scene is improved. The proposed method exploits two mechanisms, one for reducing the computational demand when calculating the eigenvectors for each matrix, and another for improving segmentation performance. The experimental results obtained with two sets of real images are presented and used to assess the proposed method. |
|---|