Metal Surface Defect Detection Using Deep Learning Techniques

Metal defect detection plays a crucial role in ensuring the quality and safety of industrial manufacturing processes. Traditional methods of defect detection often rely on manual inspection, which can be time-consuming and prone to human error. In recent years, the application of deep learning techniques, particularly Convolutional Neural Networks (CNNs), has shown promising results in automating defect detection tasks. This research explores the effectiveness of using CNN architectures, specifically InceptionNet, ResNet34, and VGG16, for automated metal defect detection. The study involves training and evaluating these models on a dataset of metal surface images containing various types of defects. The performance metrics such as accuracy, precision, recall, and F1-score are analysed to assess the robustness and efficacy of each model in defect classification. Additionally, the research investigates the impact of transfer learning by fine-tuning pre-trained CNN models on the defect detection task. The experimental results demonstrate the superior performance of deep CNNs compared to traditional methods, highlighting the potential of these techniques for real-world applications in industrial quality control and inspection.