This content is not included in your SAE MOBILUS subscription, or you are not logged in.
A Crack Detection Method for Self-Piercing Riveting Button Images through Machine Learning
ISSN: 0148-7191, e-ISSN: 2688-3627
To be published on April 14, 2020 by SAE International in United States
Self-piercing riveting (SPR) is a key joining technique for lightweight materials, and it has been widely used in the automobile manufacturing. Manual visual crack inspection of SPR joints could be time-consuming and might rely on high-level training for engineers to distinguish features subjectively. This paper presents a machine learning based crack detection method for SPR button images. Firstly, sub-images were cropped from the button images and preprocessed into three categories (cracks, edges, others) as training samples. Then, Artificial Neural Network (ANN) was chosen as the classification algorithm for sub-images. During the training of ANN, three pattern descriptors were proposed as feature extractors of sub-images respectively, and compared with validation samples. Lastly, a search algorithm was developed to extend the application of the learned model from sub-images to the original button images. The preliminary results on non-cracked and cracked button images show that the proposed crack detection method could achieve an acceptable performance.