基于映射图与界面视觉要素排序的激光全息投影人机动态交互
Human-computer Dynamic Interaction for Laser Holographic Projection Based on Mapping and Interface Visual Element Ordering on AHP and FCE
张颖1
作者信息
摘要
针对人机动态交互界面任务执行中,界面感知元素检测过程一旦失真,存在耗时较长、操作准确率较低的问题,提出基于映射图与界面视觉要素的激光全息投影人机动态交互方法.利用自组织特征映射算法生成映射图,通过迭代学习提取目标特征;结合感知元素重要度指标和人机交互界面感知强度,通过自动标注为视觉感知元素排序编号,构建人机动态交互界面优化模式,设计激光全息投影人机动态交互界面.测试结果表明,所提方法应用下,被标记的目标视觉元素特征点经过更新后,布局完整、有序,感知强度指数误差控制在1-5之间,交互任务执行耗时均在12.5s以内,操作准确率平均值为92％,解决了激光全息投影人机动态交互优化问题.
Abstract
The paper looks into the problems of time-consuming and low operational accuracy in the detection process of interface perception elements once distorted during the task execution of human-computer dynamic interaction.A laser holographic projection human-computer dynamic interaction method based on mapping and interface vision elements is proposed.A self organizing feature mapping algorithm is used to generate a mapping map and extract target features through iterative learning.Combining the importance index of perception elements and the perception intensity of human-computer interaction interfaces,an optimization mode for human-computer dynamic interaction interfaces is constructed by automatically labeling and numbering visual perception elements,and a laser holographic projection human-computer dynamic interaction interface is designed.The test results show that under the application of the proposed method,the marked visual element feature points of the target are updated,with a complete and orderly layout,a perceptual intensity index error controlled between 1-5,interactive task execution time within 12.5 seconds,and an average operating accuracy of 92％.This solves the optimization problem of laser holographic projection human-computer dynamic interaction.
关键词
激光全息投影/人机动态交互/Java语言/目标检测算法/自组织特征映射算法Key words
laser holographic projection/human-computer dynamic interaction/Java language/target detection algorithm/self-organizing feature mapping algorithm引用本文复制引用
基金项目
湖北省教育厅科研项目(B2022376)
湖北省自然科学基金(YZ200814)
出版年
2024
NETL
NSTL
万方数据