Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Jul 8;13(1):11077.
doi: 10.1038/s41598-023-38085-x.

Aluminum surface defect detection method based on a lightweight YOLOv4 network

Songsong Li #  1 , Shangrong Guo #  2 , Zhaolong Han #  2 , Chen Kou #  2 , Benchi Huang #  2 , Minghui Luan #  2
Affiliations

Aluminum surface defect detection method based on a lightweight YOLOv4 network

Songsong Li et al. Sci Rep. .

Abstract

Deep learning is currently being used to automate surface defect detection in aluminum. The common target detection models based on neural networks often have a large number of parameters and a slow detection speed, which is not conducive to real-time detection. Therefore, this paper proposes a lightweight aluminum surface defect detection model, M2-BL-YOLOv4, based on the YOLOv4 algorithm. First, in the YOLOv4 model, the complex CSPDarkNet53 backbone network was modified into an inverted residual structure, which greatly reduced the number of parameters in the model and increased the detection speed. Second, a new feature fusion network, BiFPN-Lite, is designed to improve the fusion ability of the network and further improve its detection accuracy. The final results show that the mean average precision of the improved lightweight YOLOv4 algorithm in the aluminum surface defect test set reaches 93.5%, the number of model parameters is reduced to 60% of the original, and the number of frames per second (FPS) detected is 52.99, which increases the detection speed by 30%. The efficient detection of aluminum surface defects is realized.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1
Figure 1
YOLOv4 network structure diagram.
Figure 2
Figure 2
Inverted residual block.
Figure 3
Figure 3
Feature fusion networks.
Figure 4
Figure 4
The M2-YOLOv4 network structure diagram.
Figure 5
Figure 5
The M2-BL-YOLOv4 network structure diagram.
Figure 6
Figure 6
Eight types of defects on the aluminum surface.
Figure 7
Figure 7
The AP value of defects before and after lightweighting in the backbone.
Figure 8
Figure 8
Detection results of the M2-BL-YOLOv4 model.
Figure 9
Figure 9
Detection results of the M2-BL-YOLOv4 model on the NEU-DET dataset.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Deng Y, Zhang X. Development of aluminium and aluminium alloy. Chin. J. Nonferrous Met. 2019;29:2115–2141.
    1. Wei R, Bi Y. Research on recognition technology of aluminum profile surface defects based on deep learning. Materials. 2019;12:1681–1681. doi: 10.3390/ma12101681. - DOI - PMC - PubMed
    1. Zhang J, Kang X, Ni H, Ren F. Surface defect detection of steel strips based on classification priority YOLOv3-dense network. Ironmak. Steelmak. 2020;48:547–558. doi: 10.1080/03019233.2020.1816806. - DOI
    1. Liu R, Huang M, Gao Z, Cao Z, Cao P. MSC-DNet: An efficient detector with multi-scale context for defect detection on strip steel surface. Measurement. 2023;209:112467–112467. doi: 10.1016/j.measurement.2023.112467. - DOI
    1. Xu Z, Huang D. Defect detection on aluminum surfaces based on deep learning. J. East China Norm. Univ. 2020;2020:105–114.