Development of Augmented Reality Experience System of Indoor Damage due to Earthquake


Utilization of virtual reality (VR) and augmented reality (AR) technique is useful for education of disaster. The objective of this research is to develop a fusion technique to generate AR experience video for indoor damage using three-dimensional images with calculated results. National Research Institute for Earth Science and Disaster Resilience (NIED) is developing a virtual experience system of indoor disaster damage. This system consists of a data acquisition part and a data processing and playback part. The data acquisition part can measure and provide synchronized three-dimensional images, sound and acceleration. The system was applied to the E-Defense shake table test of the 10-story reinforced concrete (RC) building and a VR experience video of indoor damage was successfully generated. This video visualized on head mounted display (HMD) can provide a virtual experience of indoor disaster under seismic excitation. A numerical analysis system called E-Simulator, which is under construction by NIED, can analyze motions and falls of indoor furniture and fixture under seismic excitation. A finite element method based on the adaptively shifted integration (ASI) – Gauss technique is used for this analysis. This method can simulate various indoor situations which are difficult to reproduce in actual experiments with high accuracy and low calculation cost using fewer elements. To achieve the objective of this research, the data of indoor damages on the 10th floor of the 10-story RC building specimen excited on the shake-table test of the E-Defense was used. The acceleration data measured on the 10th floor was used for the motion analysis of furniture placed at various locations. A system to paste the texture from several furniture images was developed, and more realistic visualized data was produced. Furthermore, a fusion technique between numerical results and three-dimensional images was developed, in which the data are synchronized by linear interpolation. The generated images are to be shown in the presentation.