基于虚拟现实的人机交互视觉目标感知仿真
Visual Goals Perception in Human-Computer Interaction Based on Virtual Reality
熊安亚 1王芹 1徐是1
作者信息
摘要
人机交互是人与计算机或其它外部设备之间的信息交换过程.在人机交互过程中,由于网络环境复杂以及不稳定等因素,直接影响视觉目标的感知效果.为解决上述问题,提出基于虚拟现实的人机交互视觉目标感知方法.利用虚拟现实技术建立安全的人机交互网络环境,并在此环境中截取视觉目标二维图像,采用Gabor小波变换法,提取视觉目标的特征向量;采用最小二乘支持向量机,建立人机交互视觉目标感知模型,通过粒子群算法确定模型相关参数;将提取的目标特征向量输入该模型中,完成人机交互视觉目标的感知.实验结果表明,上述研究方法的平均感知精度在 95%～100%之间,高于其它两种方法的测试结果,研究方法的目标感知速度和感知效果也均更优.
Abstract
Human-computer interaction is the process of information exchange between people and computers or other external devices.In the process of human-computer interaction,the complexity and instability of network can di-rectly affect the perception effect of visual objects.Therefore,a method of sensing visual objects in the process of hu-man-computer interaction is proposed based on virtual reality.Firstly,virtual reality technology was used to build a safe environment for human-computer interaction network,and then 2D image of visual object was captured in this environment.Secondly,Gabor wavelet transform method was used to extract the feature vector of visual object.Third-ly,least square support vector machine was used to construct a model of sensing visual object in human-computer in-teraction,and then relevant parameters of the model were determined by particle swarm optimization.Finally,the ex-tracted target feature vectors were input into the model,thus completing the perception of visual objects.Experimental results prove that the average perceptive accuracy of the proposed method is between 95%and 100%,which is higher than the results of the other methods.In addition,its perception speed and perception effect are also better than oth-ers.
关键词
虚拟现实技术/人机交互/视觉目标/感知方法/支持向量机Key words
Virtual reality technology/Human-computer interaction/Visual goal/Perceptive method/Support vector machine引用本文复制引用
基金项目
重庆市教委人文社会科学研究项目(22SKJD080)
出版年
2024
NETL
NSTL
万方数据