Publications
2016
1.
Alegre, Enrique; Ruiz, Jonatan; Barreiro, Joaquín; Castejón-Limas, Manuel; Hernández, LK
Automatic segmentation of the wear region in cutting tools Artículo de revista
En: 2016.
Resumen | Enlaces | BibTeX | Etiquetas: automatic segmentation, cutting tool, extended-maxima transform, machining
@article{alegre_automatic_nodate,
title = {Automatic segmentation of the wear region in cutting tools},
author = {Enrique Alegre and Jonatan Ruiz and Joaquín Barreiro and Manuel Castejón-Limas and LK Hernández},
url = {https://d1wqtxts1xzle7.cloudfront.net/44449821/Automatic_segmentation_of_the_wear_regio20160405-32089-t0uk07-libre.pdf?1459894370=&response-content-disposition=inline%3B+filename%3DAutomatic_segmentation_of_the_wear_regio.pdf&Expires=1739456394&Signature=O9SQuejVUFfkNXaJaH-Rd1mgX1dE0MfJ9xge3ri6ekr1I71KojNIII7qijC511wXXymUmYaQ-VrMMO-dSYBRVfr8cNcqykaUiNZOb~KAfrMM9ewkyHV-di20qDSZ4nLw-Ij-eFk48OorcP5L8PdZ3CsvnFrsvtnidpUszpwZL9NZo1A49e1q05cXM2SJwQOUbmkHWQDsLmJ8WOh1bawMgkzWNIKImlI5wWwqwa86g0j4~nlHw3NnFgvrDkVMPer-3vj2yWLyDkoKxjwxVSWAAotMWJPseED8RwZV6KAEU5YxWO~bL2IgSidjRBU3zvoL5Neff4wwOEnqpLKrXomX2A__&Key-Pair-Id=APKAJLOHF5GGSLRBV4ZA},
year = {2016},
date = {2016-04-05},
abstract = {This paper presents a methodology for segmenting wear regions in cutting tools using decision rules and image processing techniques. The approach adapts to different wear levels and achieved 90% accuracy on 625 samples through visual validation.},
keywords = {automatic segmentation, cutting tool, extended-maxima transform, machining},
pubstate = {published},
tppubtype = {article}
}
This paper presents a methodology for segmenting wear regions in cutting tools using decision rules and image processing techniques. The approach adapts to different wear levels and achieved 90% accuracy on 625 samples through visual validation.
2008
2.
Barreiro, Joaquín; Castejón-Limas, Manuel; Alegre, Enrique; Hernández, LK
Use of descriptors based on moments from digital images for tool wear monitoring Artículo de revista
En: International Journal of Machine Tools and Manufacture, vol. 48, no 9, pp. 1005–1013, 2008, (Publisher: Pergamon).
Resumen | Enlaces | BibTeX | Etiquetas: Cluster Analysis, cutting tool, image analysis, Wear
@article{barreiro_use_2008,
title = {Use of descriptors based on moments from digital images for tool wear monitoring},
author = {Joaquín Barreiro and Manuel Castejón-Limas and Enrique Alegre and LK Hernández},
url = {https://www.sciencedirect.com/science/article/pii/S0890695508000096},
year = {2008},
date = {2008-01-01},
urldate = {2008-01-01},
journal = {International Journal of Machine Tools and Manufacture},
volume = {48},
number = {9},
pages = {1005–1013},
abstract = {This paper addresses the overly conservative criteria for tool replacement, which leads to tools being replaced prematurely, thus increasing costs. The study explores using moments to describe tool wear images and classify the tool condition into different wear classes. Hu and Legendre descriptors were found to perform the best for wear identification. These descriptors were analyzed using a finite mixture model (mclust) to classify tools into three wear classes: low, medium, and high. Discriminant analysis techniques, including linear and quadratic methods, were used to validate the clustering results. The paper suggests that the new wear criterion, based on the probability of belonging to a wear class, can replace the traditional conservative approach, potentially reducing tool replacement costs.},
note = {Publisher: Pergamon},
keywords = {Cluster Analysis, cutting tool, image analysis, Wear},
pubstate = {published},
tppubtype = {article}
}
This paper addresses the overly conservative criteria for tool replacement, which leads to tools being replaced prematurely, thus increasing costs. The study explores using moments to describe tool wear images and classify the tool condition into different wear classes. Hu and Legendre descriptors were found to perform the best for wear identification. These descriptors were analyzed using a finite mixture model (mclust) to classify tools into three wear classes: low, medium, and high. Discriminant analysis techniques, including linear and quadratic methods, were used to validate the clustering results. The paper suggests that the new wear criterion, based on the probability of belonging to a wear class, can replace the traditional conservative approach, potentially reducing tool replacement costs.