摘要
传统的掘进机培训效率差,缺乏一种高效安全的培训方案。目前关于掘进机培训方案的研究多数仅基于虚拟现实技术,缺乏与其他技术结合使用的相关研究。为了提高学生培训质量,设计了一种具有评价系统和姿态识别系统,并且可通过物理硬件控制虚拟掘进机的培训方案。硬件控制部分主要设计了矩阵键盘和控制手柄模拟真实控制面板和操纵手柄,软件部分使用Maya对掘进机精细建模,使用C#编程语言编写逻辑脚本。姿态识别模块使用均值滤波对图像空间域降噪处理,再用帧间差分法进行目标检测,使用Open-Pose算法提取目标人物骨点信息,最终使用模板匹配法进行姿态识别。其结果表明:姿态识别、评分系统和硬件控制三者结合使用的掘进机培训方案可以有效规范学员操作姿态,提高培训质量。
Traditional boring machine training efficiency is poor,lack of an efficient and safe training program.At present,most of the researches on TBM training programs are only based on virtual reality technology,and there is a lack of research related to the combination with other technologies.In order to improve the training quality of students,a training scheme with evaluation system and attitude recognition system is designed,which can control the virtual TBM through physical hardware.In the hardware control part,matrix keyboard and control handle are mainly designed to simulate the real control panel and control handle.In the software part,Maya is used for fine modeling of the boring machine,and C#programming language is used to write logical scripts.The pose recognition module uses the mean filter to reduce noise,then uses the inter‑frame difference method to detect the target,uses the Open‑Pose algorithm to extract the bone point information of the target figure,and finally uses the template matching method to recognize the pose.The results show that the combination of attitude recognition,scoring system and hardware control can effectively standardize the trainees’attitude and improve the quality of training.
作者
谢泽浩
周明全
陈超
姚凡
谢梓炫
Xie Zehao;Zhou Mingquan;Chen Chao;Yao Fan;Xie Zixuan(School of Computer,North China Institute of Science and Technology,Langfang 065201,China)
出处
《现代计算机》
2023年第22期14-18,24,共6页
Modern Computer
关键词
隧道掘进机
虚拟仿真
姿态识别
tunnel boring machine
virtual simulation
posture recognition
