基于机器学习算法的人体步态异常状态检测与识别方法  

Research on human gait intelligent detection and recognition based on machine learning algorithm

作　　者：郭惠[1] 耿宝光 GUO Hui;GENG Baoguang(Shanxi Vocational College of Tourism,Taiyuan 030031,China;Shanxi Engineering Vocational College,Taiyuan 030031,China)

机构地区：[1]山西旅游职业学院,太原030031 [2]山西工程职业学院,太原030031

出　　处：《智能计算机与应用》2023年第10期25-28,共4页Intelligent Computer and Applications

基　　金：山西省教育科学“十四五”规划课题(GH-220321)。

摘　　要：为了提高在人体步伐识别中异常状态检测的识别率和精度,引入机器学习算法,研究人体步态异常状态检测与识别方法。根据采集到的人体步态图像序列结构,构建人体步态模型。利用所构建的模型检测人体步态,并对人体步态图像进行二值化处理,实现模型参数的优化,从中提取人体步态运动特征,实现对变化的人体步态的具体描述。结合机器学习算法中的决策树、NN分类器和KNN分类器,对人体步态中的异常状态进行识别。经过对比实验可知,本文所提识别方法具备更高的识别率和识别精度,能够有效检测出人在行走中出现的异常状态。In order to improve the recognition rate and accuracy of abnormal state detection in human gait recognition,machine learning algorithm is introduced to study the detection and recognition method of abnormal state of human gait.According to the structure of the collected human gait image sequence,the human gait model is constructed.The constructed model is used to detect human gait,and the human gait image is binarized to optimize the model parameters.The motion features of human gait are extracted to describe the changing human gait.Combining the decision tree,NN classifier and KNN classifier in machine learning algorithm,the abnormal state of human gait is recognized.The experiments prove that the recognition method proposed in this paper has higher recognition rate and recognition accuracy,and can effectively detect the abnormal state of walking.

关 键 词：机器学习算法 人体步态 智能检测 识别方法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]