Prediction from 3D Pictogram Drawing Using AR Technology for Digital Twin

Digital Twin technology, which involves creating a virtual replica using data collected from the real world, including 3D CAD and CG modeling, is an essential tool for energy conservation. The 3D direct drawing system using AR technology can create many shape models for pictograms. However, the model-creating command is difficult for a user because it is multiple and complicated. In this paper, the pictogram drawing trajectory data are investigated and accumulated. After that, their shapes are predicted for automatically converted to the desired shape using the Neural Network method, k-nearest Neighbor method, and Support Vector Machine method of machine learning, and the accuracy rates of the two methods are compared.