Detection system for correctly wearing mask based on YOLO v5 algorithm

Mengyu Liu; Rongrui Huang; Xuanyu He

doi:10.1117/12.2653657

20 October 2022 Detection system for correctly wearing mask based on YOLO v5 algorithm

Mengyu Liu, Rongrui Huang, Xuanyu He

Proceedings Volume 12350, 6th International Workshop on Advanced Algorithms and Control Engineering (IWAACE 2022); 1235012 (2022) https://doi.org/10.1117/12.2653657
Event: 6th International Workshop on Advanced Algorithms and Control Engineering (IWAACE 2022), 2022, Qingdao, China

Abstract

With the spread of the new crown epidemic and the increasing number of confirmed cases, wearing correct masks is an effective means of keeping the virus at bay. Using artificial intelligence technology to determine whether masks are being worn correctly is a possible solution. By training and learning from a certain amount of image data, target detection of masks can be achieved. RCNN or FAST-RCNN is widely used in the related field of research, although the method has good accuracy and robustness. However, there are some disadvantages such as low efficiency and long training time. In addition to this, most of the current work only stops detecting whether or not a mask is worn, but does not consider whether or not the mask is worn correctly. Therefore, this paper proposes a mask detection model based on the YOLO v5 algorithm, which uses a self-made training set to detect whether a pedestrian is wearing a mask and whether he or she is wearing it correctly. The experimental validation shows that the detection accuracy is higher, and the robustness is almost the same based on the original data set, while the practicality is greatly improved.

Citation Download Citation

Mengyu Liu, Rongrui Huang, and Xuanyu He "Detection system for correctly wearing mask based on YOLO v5 algorithm", Proc. SPIE 12350, 6th International Workshop on Advanced Algorithms and Control Engineering (IWAACE 2022), 1235012 (20 October 2022); https://doi.org/10.1117/12.2653657

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available