Mostrar el registro sencillo del ítem
Automatic visual inspection: An approach with multi-instance learning
dc.creator | Mera C. | spa |
dc.creator | Orozco-Alzate M. | spa |
dc.creator | Branch J. | spa |
dc.creator | Mery D. | spa |
dc.date.accessioned | 2017-05-12T16:05:54Z | |
dc.date.available | 2017-05-12T16:05:54Z | |
dc.date.created | 2016 | |
dc.identifier.issn | 1663615 | |
dc.identifier.uri | http://hdl.handle.net/11407/3141 | |
dc.description.abstract | One of the industrial applications of computer vision is automatic visual inspection. In the last decade, standard supervised learning methods have been used to detect defects in different kind of products. These methods are trained with a set of images where every image has to be manually segmented and labeled by experts in the application domain. These manual segmentations require a large amount of high quality delineations (on pixels), which can be time consuming and often a difficult task. Multi-instance learning (MIL), in contrast to standard supervised classifiers, avoids this task and can, therefore, be trained with weakly labeled images. In this paper, we propose an approach for the automatic visual inspection that uses MIL for defect detection. The approach has been tested with data from three artificial benchmark datasets and three real-world industrial scenarios: inspection of artificial teeth, weld defect detection and fishbone detection. Results show that the proposed approach can be used with weakly labeled images for defect detection on automatic visual inspection systems. This approach is able to increase the area under the receiver-operating characteristic curve (AUC) up to 6.3% compared with the naïve MIL approach of propagating the bag labels. © 2016 Elsevier B.V. | eng |
dc.language.iso | eng | |
dc.publisher | Elsevier B.V. | spa |
dc.relation.isversionof | http://www.sciencedirect.com/science/article/pii/S0166361516301750 | spa |
dc.source | Scopus | spa |
dc.subject | Automatic visual inspection | spa |
dc.subject | Defect detection | spa |
dc.subject | Multi-instance learning | spa |
dc.subject | Pattern recognition | spa |
dc.subject | Weak labels | spa |
dc.title | Automatic visual inspection: An approach with multi-instance learning | spa |
dc.type | Article | eng |
dc.rights.accessrights | info:eu-repo/semantics/restrictedAccess | |
dc.rights.accessrights | info:eu-repo/semantics/restrictedAccess | |
dc.contributor.affiliation | Mera, C., Universidad Nacional de Colombia, Sede Medellín, Departamento de Ciencias de la Computación y de la Decisión, Cra. 80 # 65-223, Medellín, Colombia, Universidad de Medellín, Facultad de Ingeniería, Carrera 87 # 30-65, Medellín, Colombia | spa |
dc.contributor.affiliation | Orozco-Alzate, M., Universidad Nacional de Colombia, Sede Manizales, Departamento de Informática y Computación, km 7 vía al Magdalena, Manizales, Colombia | spa |
dc.contributor.affiliation | Branch, J., Universidad Nacional de Colombia, Sede Medellín, Departamento de Ciencias de la Computación y de la Decisión, Cra. 80 # 65-223, Medellín, Colombia | spa |
dc.contributor.affiliation | Mery, D., Pontificia Universidad Católica de Chile, Departamento de Ciencias de la Computación, Av. Vicuña Mackenna 4860, Santiago de Chile, Chile | spa |
dc.identifier.doi | 10.1016/j.compind.2016.09.002 | |
dc.subject.keyword | Computer vision | eng |
dc.subject.keyword | Inspection | eng |
dc.subject.keyword | Pattern recognition | eng |
dc.subject.keyword | Automatic visual inspection | eng |
dc.subject.keyword | Automatic visual inspection systems | eng |
dc.subject.keyword | Defect detection | eng |
dc.subject.keyword | Multi-instance learning | eng |
dc.subject.keyword | Receiver operating characteristic curves | eng |
dc.subject.keyword | Supervised classifiers | eng |
dc.subject.keyword | Supervised learning methods | eng |
dc.subject.keyword | Weak labels | eng |
dc.subject.keyword | Defects | eng |
dc.relation.ispartofes | Computers in Industry | spa |
dc.type.driver | info:eu-repo/semantics/article |
Ficheros en el ítem
Ficheros | Tamaño | Formato | Ver |
---|---|---|---|
No hay ficheros asociados a este ítem. |
Este ítem aparece en la(s) siguiente(s) colección(ones)
-
Indexados Scopus [1632]