文章详情

题名 · 作者 · 摘要 · 关键词 · DOI · 参考文献

基于改进YOLOv8的复杂场景跌倒检测算法

曹萌迪1,杨梦凡2,王留毅1,王东洋1

(1.华北水利水电大学,河南 郑州 450046;2.河南水利与环境职业学院,河南 郑州 450008)

栏目:人工智能 · 第 2025 年第 5 期 · 页码:66-71 · 2096-4706(2025)05-0066-06 · 中图分类号:TP311.5 · 文献标识码:A · DOI:10.19850/j.cnki.2096-4706.2025.05.012 · 出版日期:2025-03-10

摘要

针对由于光照变化、人员密集以及人体形态被遮挡等因素导致跌倒检测精度低、实时性差的问题,文章提出了一种基于改进YOLOv8的复杂场景跌倒检测算法:SLG-YOLO。设计了SC模块并嵌入C2f模块,利用边缘检测算法来提取跌倒行为的边缘特征,在SPPF模块引入Large Separable Kernel Attention(LSKA)注意力机制,利用大型可分离卷积捕捉图像广泛的上下文信息,在增强对于重要特征关注度的同时不会增加计算复杂度;同时,采用Gather and Distribute(GD)机制,结合3D卷积的多尺度特征序列融合方法,取代原有的Neck部分,提升了模型对于复杂场景和不同尺度目标的检测精度。实验结果表明,在参数量和计算量为6.2 M和10.7 FLOPs的基础上,SLG-YOLO跌倒检测算法与YOLOv8相比,准确率提升了2.2%,召回率提升了3.0%,mAP@0.5提升了3.6%,mAP@0.5:0.95提升了3.1%。

关键词

跌倒检测;边缘检测;注意力机制;YOLOv8

参考文献

  1. [1] CHEN Y S,DU R X,LUO K T,et al. Fall Detection System Based on Real-Time Pose Estimation and SVM [C]//2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE).Nanchang:IEEE,2021:990-993.
  2. [2] REN L M,PENG Y J. Research of Fall Detection and Fall Prevention Technologies: A Systematic Review [J].IEEE Access,2019,7:77702-77722.
  3. [3] DELGADO-ESCAÑO R,CASTRO F M,CÓZAR J R,et al. A Cross-dataset Deep Learning-based Classifier for People Fall Detection and Identification [J].Computer Methods and Programs in Biomedicine,2020,184:105265.
  4. [4] GIRSHICK R. Fast R-CNN [C]//2015 IEEE International Conference on Computer Vision (ICCV).Santiago:IEEE,2015:1440-1448.
  5. [5] HE K M,GKIOXARI G,DOLLÁR P,et al. Mask R-CNN [C]//2017 IEEE International Conference on Computer Vision (ICCV).Venice:IEEE,2017:2980-2988.
  6. [6] WANG C-Y,BOCHKOVSKIY A,LIAO H-Y M. YOLOv7: Trainable Bag-of-Freebies Sets New State-of-the-Art for Real-Time Object Detectors [C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).Vancouver:IEEE,2023:7464-7475.
  7. [7] BOCHKOVSKIY A,WANG C-Y,LIAO H-Y M. YOLOv4: Optimal Speed and Accuracy of Object Detection [J/OL].arXiv:2004.10934 [cs.CV].(2020-04-23).https://arxiv.org/abs/2004.10934.
  8. [8] WANG A,CHEN H,LIU L H,et al. YOLOv10: Real-time End-to-End Object Detection [J/OL].arXiv:2405.14458 [cs.CV].(2024-05-23).https://arxiv.org/abs/2405.14458v2.
  9. [9] LIU W,ANGUELOV D,ERHAN D,et al. SSD: Single Shot MultiBox Detector [C]//Computer Vision-ECCV 2016: 14th European Conference.Amsterdam:Springer Nature,2016:21-37.
  10. [10] ZHAO Y A,LYU W Y,XU S L,et al. DETRs Beat YOLOs on Real-time Object Detection [J/OL].arXiv:2304.08069 [cs.CV].(2023-04-17).https://arxiv.org/abs/2304.08069.
  11. [11] ZHAO D,SONG T,GAO J,et al. YOLO-Fall: A Novel Convolutional Neural Network Model for Fall Detection in Open Spaces [J].IEEE Access,2024,12:26137-26149.
  12. [12] QIN Y,MIAO W Q,QIAN C. A High-precision Fall Detection Model Based on Dynamic Convolution in Complex Scenes [J].Electronics,2024,13(6):1141.
  13. [13] LAU K W,PO L-M,REHMAN Y A U. Large Separable Kernel Attention: Rethinking the Large Kernel Attention Design in CNN [J/OL].arXiv:2309.01439 [cs.CV].(2023-09-04).https://arxiv.org/abs/2309.01439?context=cs.CV.
  14. [14] LIN T-Y,DOLLÁR P,GIRSHICK R,et al. Feature Pyramid Networks for Object Detection [C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Honolulu:IEEE,2017:936-944.
  15. [15] CHARFI I,MITERAN J,DUBOIS J,et al. Definition and Performance Evaluation of a Robust SVM Based Fall Detection Solution [C]//2012 Eighth International Conference on Signal Image Technology and Internet Based Systems.Sorrento:IEEE,2012:218-224.

作者简介

曹萌迪(1999—),男,汉族,河南许昌人,硕士在读,研究方向:计算机视觉;杨梦凡(1995—),男,汉族,河南郑州人,助教,硕士,研究方向:计算机视觉;王留毅(1999—),男,汉族,河南洛阳人,硕士在读,研究方向:计算机视觉;王东洋(1998—),男,汉族,河南开封人,硕士在读,研究方向:人工智能。