[1]赵普,武一.面向社区医疗的跌倒检测算法[J].中国医学物理学杂志,2023,40(12):1486-1493.[doi:DOI:10.3969/j.issn.1005-202X.2023.12.006]
　ZHAO Pu,WU Yi.Fall detection algorithm for community healthcare[J].Chinese Journal of Medical Physics,2023,40(12):1486-1493.[doi:DOI:10.3969/j.issn.1005-202X.2023.12.006]
点击复制

面向社区医疗的跌倒检测算法()
分享到：

《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]

卷:
40卷
期数:
2023年第12期
页码:
1486-1493
栏目:
医学影像物理
出版日期:
2023-12-27

文章信息/Info

Title:
Fall detection algorithm for community healthcare
文章编号:
1005-202X（2023）12-1486-08
作者:
赵普1; 武一1; 2
1.河北工业大学电子信息工程学院， 天津 300401； 2.河北工业大学电子与通信工程国家级实验教学示范中心， 天津 300401
Author(s):
ZHAO Pu1;  WU Yi1; 2
1. School of Electronics Information Engineering, Hebei University of Technology, Tianjin 300401, China 2. National Experiment Teaching Demonstration Center for Electronic and Communication Engineering, Hebei University of Technology, Tianjin 300401, China
关键词:
跌倒检测; 非局部注意力机制; 损失函数; 社区医疗
Keywords:
Keywords: fall detection non-local attention mechanism loss function community healthcare
分类号:
R318；TP.391.4
DOI:
DOI:10.3969/j.issn.1005-202X.2023.12.006
文献标志码:
A
摘要:
为解决社区中的独居老人发生跌倒时不能及时得到救治造成二次伤害的问题，提出一种面向社区医疗的跌倒检测算法。算法具有2D卷积和3D卷积两个分支，同时进行空间特征和时序特征的提取。在3D分支引入密集连接，增强时序特征提取能力；在2D分支重新设计网络的残差块，增强空间特征提取能力。在两个分支融合部分引入非局部注意力机制，增强特征融合能力。在算法的最后融合场景信息，通过SIoU损失函数和联合损失函数监督算法检测跌倒行为。在扩充后的公开URFD数据集上进行对比实验，本文算法检测准确率为98.3%，表明所提算法对于跌倒行为有较好的表现性和鲁棒性。
Abstract:
Abstract: A fall detection algorithm for community healthcare is proposed to avoid the secondary injury caused by untimely treatment when the elder living alone falls in the community. The algorithm has two branches, namely 2D convolution and 3D convolution, which allow it can extract spatial and temporal features simultaneously. The dense connections added in the 3D branch enhance the ability to extract temporal features the residual blocks in the 2D branch are redesigned to improve the ability of spatial feature extraction and a non-local attention mechanism is introduced to the branch fusion for better feature fusion. The algorithm also takes scene information into consideration, and it is supervised by SIoU loss function and the combined loss function to realize fall detection. The experiment on the expanded public URFD dataset reveals that the proposed method has a detection accuracy of 98.3%, which verifies its performance and robustness for fall detection.

相似文献/References:

[1]郑 娱,鲍 楠,徐礼胜,等.跌倒检测系统的研究进展[J].中国医学物理学杂志,2014,31(04):5071.[doi:10.3969/j.issn.1005-202X.2014.04.021]
[2]孙颖,张吟龙,王鑫,等.面向医疗护理的视觉监控医院患者跌倒检测[J].中国医学物理学杂志,2022,39(4):436.[doi:DOI:10.3969/j.issn.1005-202X.2022.04.008]
　SUN Ying,ZHANG Yinlong,WANG Xin,et al.Medical care oriented visual surveillance of patient falls in the hospital[J].Chinese Journal of Medical Physics,2022,39(12):436.[doi:DOI:10.3969/j.issn.1005-202X.2022.04.008]

备注/Memo

备注/Memo:
【收稿日期】2023-07-07 【基金项目】国家自然科学基金（51977059）；河北省自然科学基金（E2020202042） 【作者简介】赵普，硕士，研究方向：计算机视觉，E-mail: zp_dling@163.com 【通信作者】武一，博士，教授，研究方向：智能系统控制，E-mail: wuyihbgydx@163.com
更新日期/Last Update: 2023-12-27