PREDICTIVE VIRTUAL RECONSTRUCTION OF PHYSICAL ENVIRONMENTS

摘要

A method for predictively reconstructing a physical event using augmented reality, comprising: identifying relative states of objects located in a physical area by using video analysis to analyze collected video feeds from the physical event before and after a physical event involving at least one of the objects, creating a knowledge corpus including the video analysis and the collected video feeds associated with the physical event and historical information, and capturing data, by a computing device, of the physical event area, identifying possible precursor events based on the captured data and the knowledge corpus, and generating a virtual reconstruction of the physical event using the possible precursor events, displaying, by the computer device, the generated virtual reconstruction of the predicted physical event, wherein the displayed virtual reconstruction of the predicted physical event overlays an image of the physical event area.