| In recent years,Virtual Reality(VR)has become more and more commercialized and productized.Virtual reality technology provides users with rich three-dimensional information by generating virtual environments.Virtual reality needs to simulate as much of our visual,auditory,tactile,and other feelings as possible.When people receive these sensory stimuli,they also give corresponding feedback.This real interaction process is realized by integrating hardware and software.At present,virtual shopping,virtual vacation,virtual social,virtual education,virtual medical care3 etc.have also entered the public life unconsciously.With the continuous improvement and development of virtual reality technology,it will gradually penetrate into every corner of life and change life.Every aspect of it.But at the same time,the problem is that no matter how amazing the immersive experience that virtual reality can provide to users,the discomfort such as sickness,nausea,vomiting,etc.,that is,virtual reality motion sickness,is still the bottleneck for promoting virtual reality technology.At present,there is a lack of research on virtual reality motion sickness.Therefore,it is of great significance to carry out this research.The virtual scene consists of space and objects(models).People interact directly with virtual scenes.In order to reduce the vertigo,the research starts from the perspective of virtual scene design,and visually studies the scene categories and scene details.What effect does vertigo have?Taking into account the scene type,scene complexity,scene accuracy,and scene resource issues in the scene design,the number of model patches is used as the only experimental operation index,which is intended to find the relationship between the number of model patches and the vertigo in the scene design.relationship.This research carried out exploratory experiments.Firstly,the characteristics of virtual scenes were studied.The virtual scenes were divided.Then,through 3D Max modeling,Unity3D engine built virtual scenes and used C#language programming to control the subject’s lens field of view and action path.In the experiment,the test subjects wore VR equipment to watch the virtual reality scene,and the somatosensory measurement equipment collected the three physiological indexes of brain wave,heart rate variability and skin electrical signal,and combined with the Likert-scale subjective questionnaire score to quantify the sickness feeling,and then proceeded.After data preprocessing and effective data partitioning,the relationship between model patch index and virtual motion sickness was explored by K-means clustering algorithm,and then the multi-layer perceptron algorithm was used to train the model to predict VR sickness.The results of this study can be applied to scene design and secondary development in order to achieve the purpose of assisting scene design,reducing VR sickness and optimizing scene resources. |
