基于GWO-SVM的行人跌倒检测方法
    点此下载全文
引用本文:刘善良,王士华,史宝周,姜鹏,袁勇超,李振凯,亓昭敏.基于GWO-SVM的行人跌倒检测方法[J].计算技术与自动化,2023,(1):84-90
摘要点击次数: 170
全文下载次数: 0
作者单位
刘善良,王士华,史宝周,姜鹏,袁勇超,李振凯,亓昭敏 (中国机械总院集团青岛分院有限公司,山东 青岛 266300) 
中文摘要:随着老龄化问题日趋严重,老年人的安全问题越来越受到重视,跌倒行为就是导致老年人死亡或发病的主要原因之一。基于此,提出了一种基于灰狼算法(Grey Wolf Optimizer, GWO)优化支持向量机(Support Vector Machines, SVM)的行人跌倒检测方法。选择了行人运动行为中的加速度、角速度和高度作为特征参数,构建了一种基于GWO-SVM的跌倒检测模型,使用MATLAB对模型进行训练及验证。并选择其他三种模型进行对比,使用混淆矩阵对四种模型的效果进行评估。结果表明,经灰狼算法优化的支持向量机对跌倒行为的检测精确度达到95.00%,F1-Score值同样达到95.00%。
中文关键词:跌倒检  SVM  GWO  模型
 
Pedestrian Fall Detection Method Based on GWO-SVM
Abstract:The safety of the elderly is being taken more seriously, as the problem of aging becomes increasingly serious. Failure to receive timely treatment for falls is one of the major causes of death or morbidity in elderly people. On this basis, a pedestrian fall detection method based on the Grey Wolf Optimizer (GWO) optimized Support Vector Machines (SVM) is proposed in this manuscript.The acceleration, angular velocity and height in the pedestrian motion behavior are used as characteristic parameters.A fall detection model based on GWO-SVM is constructed. and the model is trained and validated using MATLAB.Moreover, the other three models are used for comparison, and the effects of the four models are evaluated using a confusion matrix.The results show that the support vector machine optimized by the gray wolf algorithm achieves 95.00% accuracy in detecting fall behavior and 95.00% F1-Score.
keywords:fall detection  SVM  GWO  model
查看全文   查看/发表评论   下载pdf阅读器