An ensemble of fine-tuned deep learning networks for wet-blue leather segmentation / Masood Aslam in JOURNAL OF THE AMERICAN LEATHER CHEMISTS ASSOCIATION (JALCA), Vol. CXVII, N° 4 (04/2022)  

[article]  inJOURNAL OF THE AMERICAN LEATHER CHEMISTS ASSOCIATION (JALCA) > Vol. CXVII, N° 4 (04/2022) . - p. 164-170 Titre : An ensemble of fine-tuned deep learning networks for wet-blue leather segmentation Type de document : texte imprimé Auteurs : Masood Aslam, Auteur ; Tariq M. Khan, Auteur ; Syed Saud Naqvi, Auteur ; Geoff Holmes, Auteur Année de publication : 2022 Article en page(s) : p. 164-170 Langues : Américain (ame) Catégories : Automatisation Evaluation visuelle Qualité -- Contrôle Surfaces -- Analyse Wet-blue (tannage) Peau tannée au chrome (le chrome donne une couleur bleue) Index. décimale : 675.2 Préparation du cuir naturel. Tannage Résumé : As part of industrial quality control in the leather industry, it is important to segment features/defects in wet-blue leather samples. Manual inspection of leather samples is the current norm in industrial settings. To comply with the current industrial standards that advocate large-scale automation, visual inspection based leather processing is imperative. Visual inspection of wet-blue leather features is a challenging problem as the characteristics of these features can take on a variety of shapes and colour variations to constitute various normal and abnormal surface regions. The aim of this work is to automatically segment leather images to detect various features/defects along with the background through visual analysis of the surfaces. To accomplish this, a deep learning-based technique is developed that learns to segment wet-blue leather surface features. On our own curated leather images dataset, the proposed ensemble network performed well, with an F1-Score of 74 percent. Note de contenu : - LITERATURE REVIEW - THE PROPOSED METHOD : Data augmentation - Architectures - Ensembling convolutional neural networks - EXPERIMENTAL DESIGN : Dataset - Feature/defect types - Ground truth labelling - Experiment configuration - Quantitative measures - RESULTS : Quantitative comparison - Qualitative comparison - Table 1 : Comparison of the segmentation performance of our proposed method with other state-of-the-art segmentation odels for the wet-blue leather dataset - Table 2 : Class-wise comparison of the segmentation performance of state-of-the-art segmentation methods DOI : https://doi.org/10.34314/jalca.v117i4.4900 En ligne : https://drive.google.com/file/d/1eeGhr_Y_PMJEhoAwWdQfF7edVAP5PF7M/view?usp=drive [...] Format de la ressource électronique : Pdf Permalink : https://e-campus.itech.fr/pmb/opac_css/index.php?lvl=notice_display&id=37563 [article]

Réservation

Réserver ce document

Exemplaires (1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
23389	-	Périodique	Bibliothèque principale	Documentaires	Disponible

Learning to recognize irregular feature on leather surfaces / Masood Aslam in JOURNAL OF THE AMERICAN LEATHER CHEMISTS ASSOCIATION (JALCA), Vol. CXVI, N° 5 (05/2021)  

[article]  inJOURNAL OF THE AMERICAN LEATHER CHEMISTS ASSOCIATION (JALCA) > Vol. CXVI, N° 5 (05/2021) . - p. 169-178 Titre : Learning to recognize irregular feature on leather surfaces Type de document : texte imprimé Auteurs : Masood Aslam, Auteur ; Tariq M. Khan, Auteur ; Syed Saud Naqvi, Auteur ; Geoff Holmes, Auteur ; Rafea Naffa, Auteur Année de publication : 2021 Article en page(s) : p. 169-178 Note générale : Bibliogr. Langues : Américain (ame) Catégories : Cuir Evaluation visuelle Qualité -- Contrôle Surfaces -- Analyse Wet-blue (tannage) Peau tannée au chrome (le chrome donne une couleur bleue) Index. décimale : 675 Technologie du cuir et de la fourrure Résumé : As part of industrial quality control in the leather industry, it is important to identify the abnormal features in wet-blue leather samples. Manual inspection of leather samples is the current norm in industrial settings. To comply with the current industrial standards that advocate large-scale automation, visual inspection based leather processing is imperative. Visual inspection of irregular surfaces is a challenging problem as the characteristics of the abnormalities can take a variety of shape and color variations. The aim of this work is to automatically categorize leather images into normal or abnormal by visual analysis of the surfaces. To achieve this aim, a deep learning based approach is devised that learns to recognize regular and irregular leather surfaces and categorize leather images on its basis. To this end, we propose an ensemble of multiple convolutional neural networks for classifying leather images. The proposed ensemble network exhibited competitive performance obtaining 92.68% test accuracy on our own curated leather images dataset. Note de contenu : - LITERATURE REVIEW - THE PROPOSED METHOD : Data augmentation - Convolutional neural network ensembles - Setting up the CNN and training process - Global average pooling - EXPERIMENTAL DESIGN : Dataset - Performance measures - Benchmark deep learning methods - State-of-the-art methods for comparison - RESULTS : Comparison with descriptors based machine learning methods - Comparison with deep learning based methods - Class Activations Maps (CAM) - Table 1 : Comparison of models parameters with and without the GAP layer - Table 2 : Computational complexity of pooling strategies - Table 3 : Hyper-parameter values used in algorithms - Table 4 : Comparison with the state-of-the-art methods in terms of classification accuracy. T acc stands for training accuracy, val acc for validation accuracy and test acc for test accuracy - Table 5 : Comparison of methods in terms of precision, recall, F1-score and AUC - Table 6 : Comparison of models in terms of accuracy, transfer learning, batch normalization, batch size using global average pooling and image size of 500 x 375 DOI : https://doi.org/10.34314/jalca.v116i5.4291 En ligne : https://drive.google.com/file/d/1Ubev3KPms4yofj6BFf-ZgdPRP_AoZado/view?usp=drive [...] Format de la ressource électronique : Pdf Permalink : https://e-campus.itech.fr/pmb/opac_css/index.php?lvl=notice_display&id=35760 [article]

Réservation

Réserver ce document

Exemplaires (1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
22726	-	Périodique	Bibliothèque principale	Documentaires	Disponible

---

1 (1 - 2 / 2)