Anomaly Detection in Nanofibrous Materials by CNN-Based Self-Similarity

Automatic detection and localization of anomalies in nanofibrous materials help to reduce the cost of the production process and the time of the post-production visual inspection process. Amongst all the monitoring methods, those exploiting Scanning Electron Microscope (SEM) imaging are the most eff...

Full description

Bibliographic Details
Main Authors:	Paolo Napoletano, Flavio Piccoli, Raimondo Schettini
Format:	Article
Language:	English
Published:	MDPI AG 2018-01-01
Series:	Sensors
Subjects:	anomaly detection defect detection industrial quality inspection quality control convolutional neural networks, nanofibrous materials
Online Access:	http://www.mdpi.com/1424-8220/18/1/209

id	doaj-1f4f6fd094014d11af1f03cd627072bf
record_format	Article
spelling	doaj-1f4f6fd094014d11af1f03cd627072bf2020-11-25T01:30:38ZengMDPI AGSensors1424-82202018-01-0118120910.3390/s18010209s18010209Anomaly Detection in Nanofibrous Materials by CNN-Based Self-SimilarityPaolo Napoletano0Flavio Piccoli1Raimondo Schettini2Department of Computer Science, Systems and Communications, University of Milano-Bicocca, Milan 20126, ItalyDepartment of Computer Science, Systems and Communications, University of Milano-Bicocca, Milan 20126, ItalyDepartment of Computer Science, Systems and Communications, University of Milano-Bicocca, Milan 20126, ItalyAutomatic detection and localization of anomalies in nanofibrous materials help to reduce the cost of the production process and the time of the post-production visual inspection process. Amongst all the monitoring methods, those exploiting Scanning Electron Microscope (SEM) imaging are the most effective. In this paper, we propose a region-based method for the detection and localization of anomalies in SEM images, based on Convolutional Neural Networks (CNNs) and self-similarity. The method evaluates the degree of abnormality of each subregion of an image under consideration by computing a CNN-based visual similarity with respect to a dictionary of anomaly-free subregions belonging to a training set. The proposed method outperforms the state of the art.http://www.mdpi.com/1424-8220/18/1/209anomaly detectiondefect detectionindustrial quality inspectionquality controlconvolutional neural networks, nanofibrous materials
collection	DOAJ
language	English
format	Article
sources	DOAJ
author	Paolo Napoletano Flavio Piccoli Raimondo Schettini
spellingShingle	Paolo Napoletano Flavio Piccoli Raimondo Schettini Anomaly Detection in Nanofibrous Materials by CNN-Based Self-Similarity Sensors anomaly detection defect detection industrial quality inspection quality control convolutional neural networks, nanofibrous materials
author_facet	Paolo Napoletano Flavio Piccoli Raimondo Schettini
author_sort	Paolo Napoletano
title	Anomaly Detection in Nanofibrous Materials by CNN-Based Self-Similarity
title_short	Anomaly Detection in Nanofibrous Materials by CNN-Based Self-Similarity
title_full	Anomaly Detection in Nanofibrous Materials by CNN-Based Self-Similarity
title_fullStr	Anomaly Detection in Nanofibrous Materials by CNN-Based Self-Similarity
title_full_unstemmed	Anomaly Detection in Nanofibrous Materials by CNN-Based Self-Similarity
title_sort	anomaly detection in nanofibrous materials by cnn-based self-similarity
publisher	MDPI AG
series	Sensors
issn	1424-8220
publishDate	2018-01-01
description	Automatic detection and localization of anomalies in nanofibrous materials help to reduce the cost of the production process and the time of the post-production visual inspection process. Amongst all the monitoring methods, those exploiting Scanning Electron Microscope (SEM) imaging are the most effective. In this paper, we propose a region-based method for the detection and localization of anomalies in SEM images, based on Convolutional Neural Networks (CNNs) and self-similarity. The method evaluates the degree of abnormality of each subregion of an image under consideration by computing a CNN-based visual similarity with respect to a dictionary of anomaly-free subregions belonging to a training set. The proposed method outperforms the state of the art.
topic	anomaly detection defect detection industrial quality inspection quality control convolutional neural networks, nanofibrous materials
url	http://www.mdpi.com/1424-8220/18/1/209
work_keys_str_mv	AT paolonapoletano anomalydetectioninnanofibrousmaterialsbycnnbasedselfsimilarity AT flaviopiccoli anomalydetectioninnanofibrousmaterialsbycnnbasedselfsimilarity AT raimondoschettini anomalydetectioninnanofibrousmaterialsbycnnbasedselfsimilarity
_version_	1725091029724430336

Anomaly Detection in Nanofibrous Materials by CNN-Based Self-Similarity

Similar Items