首页|基于YOLOv5的多类型场景口罩检测方法与实现

基于YOLOv5的多类型场景口罩检测方法与实现

Mask Detection Method and Implementation in Multi-type Scenarios Based on YOLOv5

扫码查看
针对口罩佩戴检测中存在的面部交叉遮挡时检测困难、光线明暗目标检测精度低等问题,借助YOLOv5算法,提出了一种能够实时检测是否规范佩戴口罩的检测技术,通过实验验证了该模型在无明显遮挡,光照条件良好条件下,检测准确率较高,验证了方法的有效性,该方法可适用于动物是否佩戴口罩.通过实验给出了口罩可检测与人脸遮挡边界、光强之间的关系.
To address the problems existing in the mask-wearing detection that the detection is difficult with facial cross-occlusion and the object detection accuracy of bright and dark lights is poor,with the help of YOLOv5 algorithm,this paper proposes a detection technology that can detect whether the mask is worn with standardization in real time.Through experiments,it is verified that the model has a relatively high detection accuracy under the conditions of no obvious occlusion and good lighting,which verifies the effectiveness of the method.This method is also applicable for detecting whether animals are wearing masks.Through experiments,the relationships between mask detection and face occlusion boundaries and light intensity are given.

Deep LearningYOLOv5 algorithmObject Detectionmask detection

韩铭伟、胡斌、王孟瑜、李继、程子嘉

展开 >

北方工业大学伦敦布鲁内尔学院,北京 100144

深度学习 YOLOv5算法 目标检测 口罩检测

2024

现代信息科技
广东省电子学会

现代信息科技

ISSN:2096-4706
年,卷(期):2024.8(23)