Research on the characteristics of acrophobia in virtual altitude environment

The purpose of paper is to detect the degree of acrophobia to ensure the safety of High-altitude workers. This research simulates a virtual reality environment called `plank challenges', and uses a deep learning model to detect acrophobia level of people by analyzing the volunteers' EEG. High-altitude operation is a high-risk, high-injury operation. Workers are inevitably exposed to high-altitude environments. Workers with acrophobia have a higher probability of accidents. In this paper, `plank challenges' was constructed through VR technology and volunteers performed it in high altitudes while wearing VR equipments. Virtual reality technology allows volunteers to experiment in a safe and controlled environment. We extracted the energy features of the EEG and converted it into a two-dimensional spectral image, which preserves the spatial information of the EEG data and we use a deep convolutional network to derive features from the images. It is robust to classify the Volunteers' fear level. This program may be safely applicable to detect acrophobia level of people with the virtual environments.