本文提出了一种基于YOLOx的口罩佩戴检测方法。该方法首先对输入模块进行改进,加入双向弱光自适应网络模块,引入弱光特征,增强了模型在复杂环境下检测的鲁棒性。其次,增加了各种遮挡下的口罩佩戴检测,提高复杂环境下目标的识别精度。最终,为了测试检测效果,在佩戴口罩数据集上进行对比实验。实验结果表明,改进后的算法在实验数据集上将口罩佩戴检测mAP提高了约1.35%、达到了94.75%,而且在复杂环境和弱光环境下的检测效果也得到了较好的提升,具有较强的泛化能力。
Abstract
In the paper,a mask wearing detection system based on YOLOx is proposed.Firstly,the model improves the input module by introducing bidirectional low-light adaptive network and low-light object detection techniques,thus enhancing its robustness in complex environments.Secondly,mask wearing detection under various shelters is added to improve the target identification accuracy in complex environments.Finally,to test the effect of the model,comparative experiments are conducted based on the data set of mask-wearing.The experiment results show that the improved algorithm increases the mAP of mask-wearing recognition by about 1.35%,reaching 94.75%, according to the data set.Moreover,the detection effect in complex,blurred,and low-light environments gets much better,featuring strong generalization ability.
关键词
YOLOx /
目标检测 /
口罩佩戴检测 /
弱光适应
Key words
YOLOx /
target detection /
mask wearing detection /
weak light adaptation
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] Girshick R,Donahue J,Darrell T,et al.Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation[J].IEEE Computer Society,2013.
[2] Girshick R.Fast R-CNN[J].Computer Science,2015.
[3] Ren S,He K,Girshick R,et al.Faster R-CNN:Towards Real-Time Object Detection with Region Proposal Networks[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2017,39(6):1137-1149.
[4] Liu W,Anguelov D,Erhan D,et al.SSD:Single Shot MultiBox Detector[J].Springer,Cham,2016.
[5] Redmon J,Farhadi A.YOLOv3:An Incremental Improvement[J].arXiv e-prints,2018.
[6] Bochkovskiy A,Wang C Y,Liao H.YOLOv4:Optimal Speed and Accuracy of Object Detection[J].Computer Science,2020(4).
[7] 叶茂,马杰,王倩,等.多尺度特征融合的轻量化口罩佩戴检测算法[J].计算机工程,2021(10):1-13.
[8] 朱杰,王建立,王斌.基于YOLOv4-tiny改进的轻量级口罩检测算法[J].液晶与显示,2021,36(11):1525-1534.
[9] 薛均晓,程君进,张其斌,等.改进轻量级卷积神经网络的复杂场景口罩佩戴检测方法[J].计算机辅助设计与图形学学报,2021,33(7):1045-1054.
[10] 牛作东,覃涛,李捍东,等.改进RetinaFace的自然场景口罩佩戴检测算法[J].计算机工程与应用,2020,56(12):1-7.
[11] 于硕,李慧,桂方俊,等.复杂场景下基于YOLOv5的口罩佩戴实时检测算法研究[J].计算机测量与控制,2021,29(12):188-194.
[12] 焦双健,孙萌雪.基于深度学习的口罩佩戴状态目标检测系统设计[J].单片机与嵌入式系统应用,2021,21(11):5-9.
[13] Ge Z,Liu S,Wang F,et al.YOLOX: Exceeding YOLO Series in 2021[J]. 2021.
[14] Guo C,Li C,J Guo,et al. Zero-Reference Deep Curve Estimation for Low-Light Image Enhancement[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR),2020.
[15] Isola P,Zhu J Y,Zhou T,et al. Image-to-Image Translation with Conditional Adversarial Networks[J].IEEE,2016.
[16] Wang W,Yang W,Liu J.HLA-Face:Joint High-Low Adaptation for Low Light Face Detection[J].2021.
PDF(3788 KB)
