Autor: |
Malarvizhi, K., Lebo, J. Clifford, Kirubakaran, G., Selva, V. |
Předmět: |
|
Zdroj: |
Grenze International Journal of Engineering & Technology (GIJET); 2023, Vol. 9 Issue 2, p877-884, 8p |
Abstrakt: |
Defects in fabrics affect the fabric quality at a larger scope. The goal of this project is to create sophisticated machine learning methods for computer-aided defect detection and classification. Deep convolutional neural networks are built to learn from different defect datasets during the training phase. The model was trained using customized YoloV5 model which archives a better real time performance. On detection of the defects they are documented separately and further analyzed to provide an ideology about the various defects using the deep learning algorithms. The defect classification accuracy is improved in the proposed model. The most important performance metrics of customized CNN models are classification accuracy, model complexity, and training time. Applying the defect detection methods improves product quality, meets customer requirements, and reduces cost depreciation. [ABSTRACT FROM AUTHOR] |
Databáze: |
Complementary Index |
Externí odkaz: |
|