Abstract
Aluminum profile surface defects can greatly affect the performance, safety, and reliability of products. Traditional human-based visual inspection has low accuracy and is time consuming, and machine vision-based methods depend on hand-crafted features that need to be carefully designed and lack robustness. To recognize the multiple types of defects with various size on aluminum profiles, a multiscale defect-detection network based on deep learning is proposed. Then, the network is trained and evaluated using aluminum profile surface defects images. Results show 84.6%, 48.5%, 96.9%, 97.9%, 96.9%, 42.5%, 47.2%, 100%, 100%, and 43.3% average precision (AP) for the 10 defect categories, respectively, with a mean AP of 75.8%, which illustrate the effectiveness of the network in aluminum profile surface defects detection. In addition, saliency maps also show the feasibility of the proposed network.
Highlights
Aluminum alloys have drawn more and more attention in aerospace engineering, automotive, and electronics industries due to their low density, high specific strength, good corrosion resistance, and good recycling ability [1,2]
Asea Brown Boveri (ABB) Metallurgy [4] developed the Decraktor detection unit P1 using the principle of multi-frequency eddy current testing
In the training of the network, we proposed the use of one in batch size, and 50 epochs to improve the performance
Summary
Aluminum alloys have drawn more and more attention in aerospace engineering, automotive, and electronics industries due to their low density, high specific strength, good corrosion resistance, and good recycling ability [1,2]. The surface quality of aluminum profiles has assumed significant importance. Any surface defects, such as cracks and deformations, will greatly affect the performance, safety, and reliability of products. Human-based visual inspection is a common detection method in manufacturing engineering. Due to low sampling rate, low precision, poor real-time performance, fatigue, greatly influenced by artificial experience, and other adverse factors, the artificial inspection is not sufficient to guarantee the stability and accuracy of detection. The other methods based on various signals, like electrical signal and magnetic signal, were utilized to detect the surface defects by many companies. Eddy current testing can only detect conductors and needs to be close to the surface being inspected. The rough surface affects the detection result and the penetration depth of eddy current detector is limited
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
