机械制造与自动化2025，Vol.54Issue(2)：17-21,31,6.DOI:10.19344/j.cnki.issn1671-5276.2025.02.003
基于3D视觉感知的警务实战训练与评价方法研究
Research on Police Actual Combat Training and Evaluation Methods Based on 3D Visual Perception
摘要
Abstract
This paper proposes a police tactical training system integrating 3D visual motion capture and virtual reality technologies to enhance the effectiveness of physical fitness and skill development.The system employs multimodal data fusion technology to acquire trainees'kinematic characteristics in real time,establishing a standardized motion feature database.Dynamic matching between trainee actions and reference templates is achieved with temporal alignment algorithm,and analysis on motion deviation patterns are complemented by hierarchical machine learning models.With the innovative corporation of adaptive feedback mechanism,reinforcement learning strategies are utilized to optimize instructional protocols dynamically,forming a closed-loop training paradigm of"assessment-feedback-reinforcement".The proposed system significantly improves the scientific rigor and adaptability of police tactical training,providing an innovative solution for intelligent training of actual combat competency.关键词
3D视觉动作捕捉/虚拟现实/多模态数据融合/警务实战训练Key words
3D visual motion capture/virtual reality/multi-modal data fusion/police combat training分类
信息技术与安全科学引用本文复制引用
刘涛,史曙生,刘俊峰,蒋天宇,刘凯,姚佳烽..基于3D视觉感知的警务实战训练与评价方法研究[J].机械制造与自动化,2025,54(2):17-21,31,6.基金项目
2021年国家社科基金项目(21BTY024) （21BTY024）
江苏警官学院2024年教学改革重点项目(024A01) （024A01）
