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
. 2022 Sep 5;22(17):6702.
doi: 10.3390/s22176702.

Workshop Safety Helmet Wearing Detection Model Based on SCM-YOLO

Bin Zhang  1 , Chuan-Feng Sun  1 , Shu-Qi Fang  1 , Ye-Hai Zhao  2 , Song Su  2
Affiliations

Workshop Safety Helmet Wearing Detection Model Based on SCM-YOLO

Bin Zhang et al. Sensors (Basel). .

Abstract

In order to overcome the problems of object detection in complex scenes based on the YOLOv4-tiny algorithm, such as insufficient feature extraction, low accuracy, and low recall rate, an improved YOLOv4-tiny safety helmet-wearing detection algorithm SCM-YOLO is proposed. Firstly, the Spatial Pyramid Pooling (SPP) structure is added after the backbone network of the YOLOv4-tiny model to improve its adaptability of different scale features and increase its effective features extraction capability. Secondly, Convolutional Block Attention Module (CBAM), Mish activation function, K-Means++ clustering algorithm, label smoothing, and Mosaic data enhancement are introduced to improve the detection accuracy of small objects while ensuring the detection speed. After a large number of experiments, the proposed SCM-YOLO algorithm achieves a mAP of 93.19%, which is 4.76% higher than the YOLOv4-tiny algorithm. Its inference speed reaches 22.9FPS (GeForce GTX 1050Ti), which meets the needs of the real-time and accurate detection of safety helmets in complex scenes.

Keywords: K-Means++ clustering algorithm; YOLOv4-tiny; convolutional block attention module; label smoothing; safety helmet wearing detection; spatial pyramid pooling structure.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1
Figure 1
YOLOv4-tiny network structure.
Figure 2
Figure 2
CSPNet structure.
Figure 3
Figure 3
SCM-YOLO network structure.
Figure 4
Figure 4
Comparison of the activation functions.
Figure 5
Figure 5
The original SPP structure.
Figure 6
Figure 6
SPP structure in this paper.
Figure 7
Figure 7
CBAM structure.
Figure 8
Figure 8
Channel attention module structure.
Figure 9
Figure 9
Spatial attention module structure.
Figure 10
Figure 10
Image preprocessing.
Figure 11
Figure 11
Mosaic data enhancement.
Figure 12
Figure 12
K-Means++ cluster center map. The dots with different colors represent anchor boxes of different sizes. There are six colors in the figure, representing six sizes of anchor boxes. The six symbols “+” represent the cluster centers of the six anchor boxes, respectively.
Figure 13
Figure 13
Part of the dataset and LabelMe annotations.
Figure 14
Figure 14
Comparison of the loss functions.
Figure 15
Figure 15
Comparative experimental results.
Figure 16
Figure 16
AP curves for head and helmet.
Figure 17
Figure 17
CBAM visualization experiment results.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Le N., Rathour V.S., Yamazaki K., Luu K., Savvides M. Deep reinforcement learning in computer vision: A comprehensive survey. Artif. Intell. Rev. 2021;55:2733–2819. doi: 10.1007/s10462-021-10061-9. - DOI
    1. Campero-Jurado I., Márquez-Sánchez S., Quintanar-Gómez J., Rodríguez S., Corchado J.M. Smart helmet 5.0 for industrial internet of things using artificial intelligence. Sensors. 2020;20:6241. doi: 10.3390/s20216241. - DOI - PMC - PubMed
    1. Otgonbold M.-E., Gochoo M., Alnajjar F., Ali L., Tan T.-H., Hsieh J.-W., Chen P.-Y. SHEL5K: An extended dataset and benchmarking for safety helmet detection. Sensors. 2022;22:2315. doi: 10.3390/s22062315. - DOI - PMC - PubMed
    1. Yue S., Zhang Q., Shao D., Fan Y., Bai J. Safety helmet wearing status detection based on improved boosted random ferns. Multimed. Tools Appl. 2022;81:16783–16796. doi: 10.1007/s11042-022-12014-y. - DOI
    1. Gu Y., Wang Y., Shi L., Li N., Zhuang L., Xu S. Automatic detection of safety helmet wearing based on head region location. IET Image Process. 2021;15:2441–2453. doi: 10.1049/ipr2.12231. - DOI

LinkOut - more resources