基于穿戴式设备的狭小空间增强现实装配引导技术
Assembly Guidance Technology for Augmented Reality in Narrow Spaces Based on Wearable Devices
雷颖杰 1刘佳兴 2刘省委2
作者信息
- 1. 中国兵器装备集团兵器装备研究所,北京 102200;西北工业大学机电学院,西安 710000
- 2. 中国兵器装备集团兵器装备研究所,北京 102200
- 折叠
摘要
装配效率是制约我国先进战机产量提升的关键,存在装配区域狭小,工人无法看到零件与装配区域相对位置,仅依靠触觉反馈进行装配的问题.提出了面向狭小空间的增强现实装配引导技术,利用基于FAST检测及SIFT描述的方法进行特征点检测,通过改进Census变换的双目匹配方法,实现特征点匹配.实验结果表明,可穿戴式增强现实装配引导方法相比于传统装配方式效率提升24.75％,准确率由95％提升至99％,该智能化方法可显著提升狭小区域零件装配效率与质量.
Abstract
Assembly efficiency is the key factor restricting the increase in production of advanced fighter jets in China.There is a problem of narrow assembly areas,where workers cannot see the relative positions of parts and assembly areas,and rely solely on tactile feedback for assembly.A augmented reality assembly guidance technology for narrow spaces is proposed,the methods based on FAST detection and SIFT description for feature point detection are utilized.By improving the binocular matching method of Census transformation,feature point matching is achieved.The experimental results show that the wearable augmented reality assembly guidance method has an efficiency improvement of 24.75％compared with the traditional assembly methods,and the accuracy has been improved from 95％to 99％.This intelligent method can significantly improve the efficiency and quality of narrow area component assembly.
关键词
狭小空间/零件装配/增强现实/双目视觉/穿戴式设备Key words
narrow spaces/component assembly/augmented reality/binocular vision/wearable devices引用本文复制引用
基金项目
国家重点研发计划项目(2021YFB1714900)
国家自然科学基金(52275513)
出版年
2024
NETL
NSTL
万方数据