摘要
特征选择和分类器参数优化是提高人体行为识别率的关键技术,针对当前模型没有考虑两者之间的联系不足,为了提高人体行为的识别率,提出了一种特征选择和分类器参数优化联合进行的人体行为识别模型。首先,分析当前人体行为识别研究的现状,并建立人体行为识别特征和分类器参数优化的数学模型;然后,利用改进粒子群算法对数学模型进行求解,建立最优的人体行为识别模型;最后,通过仿真实验测试其性能。结果表明,其模型克服了人体行为识别模型的缺陷,提高了人体行为识别率,识别速度也要快于对比模型。
Feature selection and classifier parameter optimization are key techniques to improve the recognition rate of human behavior,current models do not consider link between feature selection and classifier parameter optimization.In order to improve the human behavior recognition rate,a feature selection and classifier parameter optimization model for human behavior recognition is proposed.First of all,it analyzes the current situation of research on human behavior recognition,and a mathematical model for optimization of human behavior recognition features and classifier parameters is established.Secondly,improved particle swarm optimization algorithm is used to solve mathematical model,and the optimal human behavior recognition model is established.Finally,the performance is tested by simulation experiments.The results show that the model overcomes the defects of the current human behavior recognition model,and improves the recognition rate of human behavior,and the recognition speed is faster than the contrast model.
出处
《微型电脑应用》
2016年第4期74-77,共4页
Microcomputer Applications
关键词
人体行为
特征选择
分类器参数优化
粒子群算法
Human Behavior
Feature Selection
Parameter Optimization
Particle Swarm Optimization
