This research work focuses on using deep learning models, ResNet50 and MobileNet architectures to classify tyres into defective and good categories. Tyre problem identification is a necessary task for which machine learning may significantly increase accuracy and efficiency, therefore ensuring vehicle safety. MobileNet far outperforms ResNet50 based on an accuracy of $\mathbf{9 5. 6 8 \%}$ against $\mathbf{7 6. 0 6 \%}$. MobileNet’s lightweight architecture with depthwise separable convolutions allows it to manage binary classification jobs more effectively, especially in environments with limited processing resources or fewer datasets. MobileNet is highlighted in this work as a better option for real-time industrial applications, particularly in automated tyre inspection systems where fast processing and exceptional accuracy are essential. This work with a thorough study of the application of transfer learning models for tyre quality classification as well as substantial new avenues for smart manufacturing and vehicle care systems. The effective classification gives a detailed point to determine the quality of the tyre as good in shape and to check any deformity in the tyres to enhance the working of the automobile.
Enhancing Tyre Quality Assessment with Cutting-Edge Classification Algorithms
06.11.2024
885583 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Engineering Index Backfile | 1950
Classification of aerodynamic tyre characteristics
Tema Archiv | 2014
|Enhancing brush tyre model accuracy through friction measurements
Taylor & Francis Verlag | 2022
|Tyre Yarn to Tyre Cord at High Quality and Low Cost
British Library Conference Proceedings | 2000
|