Publications
2011
1.
Suárez-Castrillón, Alexci; Barreiro, Joaquín; Alegre, Enrique; García-Ordás, María Teresa; Olivera, Óscar García-Olalla
MATERIAL SURFACE CHARACTERIZATION USING LAWS DESCRIPTORS. Artículo de revista
En: Annals of DAAAM & Proceedings, 2011.
Resumen | Enlaces | BibTeX | Etiquetas: Linear Discriminant Analysis, Material Classification, texture descriptors, Vision System
@article{suarez_castrillon_material_2011,
title = {MATERIAL SURFACE CHARACTERIZATION USING LAWS DESCRIPTORS.},
author = {Alexci Suárez-Castrillón and Joaquín Barreiro and Enrique Alegre and María Teresa García-Ordás and Óscar García-Olalla Olivera},
url = {https://scholar.google.es/citations?view_op=view_citation&hl=es&user=opCbArQAAAAJ&cstart=100&pagesize=100&sortby=title&citation_for_view=opCbArQAAAAJ:R3hNpaxXUhUC},
year = {2011},
date = {2011-01-01},
urldate = {2011-01-01},
journal = {Annals of DAAAM & Proceedings},
abstract = {A vision system is presented that uses texture descriptors to classify materials in industrial processes. The system was tested on five materials: concrete, sandstone, wood, metal, and paper. Haralick descriptors from the co-occurrence matrix and first-order descriptors combined with Laws' energy were used. Classification was performed using linear discriminant analysis. The results demonstrate that the R5R5 Laws mask provides better classification performance, with higher hit rates compared to the co-occurrence matrix.},
keywords = {Linear Discriminant Analysis, Material Classification, texture descriptors, Vision System},
pubstate = {published},
tppubtype = {article}
}
A vision system is presented that uses texture descriptors to classify materials in industrial processes. The system was tested on five materials: concrete, sandstone, wood, metal, and paper. Haralick descriptors from the co-occurrence matrix and first-order descriptors combined with Laws' energy were used. Classification was performed using linear discriminant analysis. The results demonstrate that the R5R5 Laws mask provides better classification performance, with higher hit rates compared to the co-occurrence matrix.