Publications

@article{riego_strong_2021,

title = {Strong classification system for wear identification on milling processes using computer vision and ensemble learning},

author = {Virginia Riego and Manuel Castejón-Limas and Lidia Sánchez-González and Laura Fernández-Robles and Hilde Perez and Javier Díez-González and Ángel-Manuel Guerrero-Higueras},

url = {https://www.sciencedirect.com/science/article/pii/S0925231220316155},

year  = {2021},

date = {2021-01-01},

journal = {Neurocomputing},

volume = {456},

pages = {678–684},

abstract = {This paper proposes a machine-vision-based method for evaluating the texture of the inner and outer surfaces of machined metallic pieces with cylindrical holes. Using a microscope camera connected to a rigid industrial boroscope, images of the hole surface are captured. The texture descriptors extracted from these images are analyzed, and a significant correlation is found. Feature reduction is performed, followed by classification using various algorithms with exhaustive grid search and 10-fold cross-validation. The best results are obtained with the Extremely Randomized Trees classifier, achieving a mean test score of 92.98%, surpassing previous research and meeting industry requirements.},

note = {Publisher: Elsevier},

keywords = {Haralick Descriptors, Milling Machined Parts, Quality Estimation, Wear Detection},

pubstate = {published},

tppubtype = {article}

}

@article{castejon-limas_texture_2019,

title = {Texture descriptors for automatic estimation of workpiece quality in milling},

author = {Manuel Castejón-Limas and Lidia Sánchez-González and Javier Díez-González and Laura Fernández-Robles and Virginia Riego and Hilde Pérez},

url = {https://link.springer.com/chapter/10.1007/978-3-030-29859-3_62},

year  = {2019},

date = {2019-01-01},

journal = {Hybrid Artificial Intelligent Systems: 14th International Conference, HAIS 2019, León, Spain, September 4–6, 2019, Proceedings 14},

pages = {734–744},

abstract = {This paper presents a method for automatically detecting workpiece wear in milling using texture descriptors. Images of the inner surface of the workpiece are captured with a boroscope and analyzed. Texture features are computed from the co-occurrence matrix, and feature vectors are classified using four methods: Decision Trees, K Neighbors, Naïve Bayes, and Multilayer Perceptron. Linear discriminant analysis reduces the feature set from six to two without sacrificing accuracy. The Decision Trees approach achieves a 91.8% hit rate, meeting industrial requirements.},

note = {Publisher: Springer International Publishing},

keywords = {Haralick Descriptors, Milling Machined Parts, Quality Estimation, Wear Detection},

pubstate = {published},

tppubtype = {article}

}

@article{alegre_surface_2010,

title = {Surface finish control in machining processes using haralick descriptors and neuronal networks},

author = {Enrique Alegre and Rocío Alaiz-Rodríguez and Joaquín Barreiro and Eduardo Fidalgo and Laura Fernández-Robles},

url = {https://link.springer.com/chapter/10.1007/978-3-642-12712-0_21},

year  = {2010},

date = {2010-01-01},

journal = {Computational Modeling of Objects Represented in Images: Second International Symposium, CompIMAGE 2010, Buffalo, NY, USA, May 5-7, 2010. Proceedings 2},

pages = {231–241},

abstract = {This paper presents a computer vision-based method to control surface roughness in steel parts. It classifies steel surfaces into acceptable and defective classes based on roughness. The study uses 143 images of AISI 303 stainless steel and three image description methods: texture local filters, Haralick descriptors, and wavelet transform features. The best error rate of 4.0% was achieved using texture descriptors with K-NN, while the optimal configuration with a neural network achieved a 0.0% error rate using Haralick descriptors.},

note = {Publisher: Springer Berlin Heidelberg},

keywords = {Classification Methods, Computer vision, Haralick Descriptors, Surface Finish Control, surface roughness},

pubstate = {published},

tppubtype = {article}

}